A strategy to enhance CO2 permeability of well-defined hyper-branched polymers with dense polyoxyethylene comb graft

Author’s Accepted Manuscript A strategy to enhance CO 2 permeability of welldefined hyper-branched polymers with dense polyoxyethylene comb graft Ikuo Taniguchi, Norihisa Wada, Kae Kinugasa, Mitsuru Higa www.elsevier.com/locate/memsci

PII: DOI: Reference:

S0376-7388(17)30509-4 http://dx.doi.org/10.1016/j.memsci.2017.04.046 MEMSCI15210

To appear in: Journal of Membrane Science Received date: 20 February 2017 Revised date: 12 April 2017 Accepted date: 21 April 2017 Cite this article as: Ikuo Taniguchi, Norihisa Wada, Kae Kinugasa and Mitsuru Higa, A strategy to enhance CO 2 permeability of well-defined hyper-branched polymers with dense polyoxyethylene comb graft, Journal of Membrane Science, http://dx.doi.org/10.1016/j.memsci.2017.04.046 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. 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.

A strategy to enhance CO2 permeability of well-defined hyper-branched polymers with dense polyoxyethylene comb graft Ikuo Taniguchi1,*, Norihisa Wada2, Kae Kinugasa1, Mitsuru Higa2,* 1

International Institute for Carbon-Neutral Energy Research (WPI-I2CNER), Kyushu University, 744

Motooka, Nishi-ku, Fukuoka 819-0395, Japan 2

Graduate School of Science and Engineering, Yamaguchi University, 2-16-1 Tokiwadai, Ube

755-8611, Japan [email protected] [email protected] *

Correspondence author: Tel: +81 92 802 6712; Fax: +81 92 802 6712.

*

Correspondence author. Tel: +81 836 85 9203; Fax: +81 836 85 9201.

Abstract Hyper-branched polymers comprised of dense polyoxyethylene (POE) comb graft and poly(methyl methacrylate) (PMMA) backbone are prepared with a well-defined chemical structure and the gas transportation properties are investigated. The CO2 permeation is strongly dependent on the POE weight fraction. To enhance the CO2 permeability, a thin film layer of the hyper-branched polymers is formed on a polydimethylsiloxane (PDMS) support with thickness less than 60 µm, where poly(vinyl alcohol) (PVA) is blended to improve the membrane formability. The CO2 permeability is increased by decreasing the thickness of the CO2-selective layer. The permeability coefficient of the resulting thin film composite (TFC) membranes exceeds 1,000 barrer with 40.5 wt% of POE methacrylate (POEM) fraction, when the thickness of the selective layer is smaller than 25 m. It reaches a maximum of 1,470 barrer for 15 m selective layer thickness with a CO2/N2 selectivity of 24.5. The permeability coefficient of the selective layer alone is 490 barrer, with CO2 selectivity of >30. Formation of a POE-rich domain upon microphase separation is confirmed by DSC and SAXS, and this is deemed crucial to enhance CO2 permeability, due to improved CO2 solubility in the selective layer. A dense POE comb architecture on the graft chain results in higher CO2 permeability than that on the polymer backbone.

1

Keywords: CO2 separation membrane; hyper-branched structure; microphase separation; poly(oxyethylene); thin film composite membrane Introduction

Precise design of chemical and molecular structures is important for clarifying the interplay between the material architecture and properties. For polymeric materials, a well-defined chemical structure governs the morphology, which in turn can also determine the functions and/or properties. This relationship is often observed in naturally occurring biopolymers such as proteins, and synthetic block copolymers (BCPs). In particular, self-assembly of BCPs upon microphase separation has been studied intensively [1,2], and the resulting nanoarchitectures have been utilized for biomedical applications [3,4], various nanotechnologies including nanolithography [5,6], and separation membranes [7]. POE-containing block copolymers (for example Pebax, a linear multiblock of POE and poly(ester-amide) segments) have been developed as membrane materials for CO2 capture [8-10] due to the inherent CO2-philic nature of POE [11]. Among the POE-containing polymers, POE is mostly incorporated into the polymer backbone [8-10, 12-20], which reduces flexibility of the POE block in comparison to tethering onto a polymer backbone, which may cause a decrease of CO2 diffusion. In addition, POEs with molecular weight over 2,000 display increased crystallinity, which also decreases the gas diffusivity [20]. To date it has been challenging to introduce amorphous or low molecular weight POE with higher weight fraction into polymer matrices. One successful example was a cross-linked POE network by Freeman and coworkers [11, 21-26], prepared by photopolymerization of poly(ethylene glycol) diacrylate (PEGDA)/poly(ethylene glycol) methyl ether acrylate (PEGMEA). The highest POE fraction was 82 wt% in the polymeric membranes, and this strategy improved the fractional free volume for gas permeation by suppressing crystallization of POE. The obtained membranes exhibited excellent CO2 separation properties even over smaller H2 [24]. In addition, the CO2 transport properties can be tuned by changing the PEGDA/PEGMEA ratio or crosslinking density.

2

We developed hyper-branched polymers, with a dense POE comb tether on a PMMA or poly(styrene-r-acrylonitrile) backbone, which controlled graft chain distance along the polymer backbone, and controlled comb length / degree of polymerization of POEM, as shown in Fig. 1 [27]. In comparison to block copolymer synthesis, tethering polymeric blocks would be more feasible. Dense POE comb grafting was conducted on PVC to form PVC-g-POEM for polymer electrolytes [28]. The POE weight fraction was up to 56 wt% having 5 ethylene glycol units with methoxy terminal groups (POEM: ~78 wt%) in the PMMA-g-PPOEM hyper-branched polymer (where PPOEM stands for polyPOEM). In comparison to POE-containing linear BCPs and cross-linked POEs as described above, the dense comb structure with shorter POE was more amorphous, and expected to have higher chain flexibility. Our previous study revealed that hyper-branched polymers demonstrated high CO2 permeability with good CO2 selectivity [27]. In addition, higher POEM incorporation gave rise to not only increased solubility, but also increased diffusivity in the polymeric membranes. For example, a self-standing membrane of PMMA-g-PPOEM hyper-branched polymers (200 m in thickness), resulted in a maximum CO2 permeability of 101 barrer (1 barrer=10-10 cm3(STP)cm/(cm2·s·cmHg)) with CO2/N2 selectivity of 38, for a POE fraction of 49 wt%.

Fig. 1. Synthetic pathways of PMMA-g-PPOEM hyper-branched polymer (above) and

3

PMMA-b-PPOEM block copolymer (below).

While various POE containing polymers (including hyper-branched polymers) have been prepared, there are still common problems to be solved for use as CO2 separation membrane materials. The CO2 transport properties should be enhanced, especially in post-combustion CO2 capture at e.g. coal-fired plants, where CO2 is separated from N2 [29,30]. One plausible approach to enhance the gas permeability is reducing the membrane thicknesses. However, for hyper-branched polymers, membranes cannot generally be self-standing nor mechanically robust after reducing the membrane thickness to less than e.g. 80 m. In this report, we investigate thin film layer formation of PMMA-g-PPOEM hyper-branched polymers on a commercial PDMS support by spin-coating. This results in a thin film composite (TFC) membrane, where PMMA-g-PPOEM acts as a CO2 selective layer with the thickness below 60 µm. In the thin layer preparation process, PVA is blended with the hyper-branched polymers to improve membrane formability and mechanical properties and to suppress formation of major pinhole defects, which is often seen when reducing film thickness. Our previous study also indicated that the POE weight fraction, or presence of a POE-rich domain, is a key factor in enhancing the CO2 separation performance of polymeric membranes [27], although the detailed morphology of the POE-rich domain remains unknown. Thus, further morphological analysis is carried out by small-angle x-ray scattering (SAXS) to clarify the interplay between the gas transport properties and membrane morphology of the CO2-selective layer.

Experimental

1. Materials Monomers, methyl methacrylate (MMA), 4-chloromethyl styrene (Cl-St), POE methacrylate (POEM, average Mn: 500, average oxyethylene unit: 9), radical initiator, ethyl 2-bromoisobutyrate (EBI), and PVA (Mw: 85-124 kDa, 99+% hydrolyzed) were obtained from Sigma-Aldrich (MO,

4

USA). The radical inhibitor in the monomers was removed by passing through a basic alumina column (MP Biomedicals, Eschwege, Germany) immediately before use. Copper(I) chloride (CuCl), 2,2’-azobisisobutyronitrile (AIBN), and 4,4’-dimethyl-2,2’-dipyridyl (bpy) were purchased from Wako (Tokyo, Japan). AIBN was purified by recrystallization from methanol to obtain needle-like crystals. Other organic and inorganic chemicals were reagent grade and used without further purification. A PDMS sheet with 62 m thickness was available from As One (Cat.# 3-345-095, Osaka, Japan) and used as a high gas permeable support for TFC membranes.

2. Preparation of hyper-branched polymer Hyper-branched polymers were chemically synthesized using the same procedures as previously reported [27], except for POEM with a longer POE chain to prepare dense comb graft on the PMMA backbone. While the average oxyethylene repeating unit of POEM was 5 in our previous report, it was increased to 9 in this research. The longer POE is expected to result in ready formation of a POE-rich domain upon microphase separation, through which CO2 would migrate preferentially. The longer POE chain does not show crystalline nature in this case, which is known to suppress gas permeation [20]. Hyper-branched polymers were obtained by the following 2-step polymerization process. First, PMMA macroinitiator was prepared by free radical copolymerization of MMA and Cl-St with AIBN. Detailed synthetic conditions and results are summarized in Table S1 in the supporting information. According to these procedures, macroinitiators with various Cl-St fractions were obtained with Mw over 50 kDa, which was high enough to fabricate polymeric membranes. In addition, when the molar fraction of Cl-St is smaller than 2 % in the resulting polymers, the Cl moiety is distributed statistically along the PMMA macroinitiators, in other words, the average distance between the moieties can be controlled. A dense POEM comb was then propagated from the Cl moieties by atom transfer radical polymerization (ATRP) to form a hyper-branched structure. The ATRP has been employed for living polymerization of various monomers including methacrylates [31-33], and POEM graft chains with narrow molecular weight distribution would result from the polymerization technique. In brief, the macroinitiator was dissolved in N-methylpyrrolidone (NMP), and predetermined amount of

5

POEM, CuCl, and bpy were added to the solution. Here, the POE monomer, CuCl, and the ligand were 1:1.2:2.4 by the molar ratio. The ATRP was conducted at 90°C for 24-72 h after N2 bubbling for 30 min at ambient temperature to eliminate molecular oxygen in the reaction mixture. The obtained product was reprecipitated in petroleum ether/ethanol (4:1 by vol.) with trace amount of HCl to remove copper from the active chain end on a crushed ice bath. The precipitate was dissolved in toluene, and the solution was passed through a silica gel column to remove the transition metal. The hyper-branched polymers were further purified by reprecipitation in petroleum ether twice. Introduction of the POEM comb was confirmed by 1H NMR on a JNM-EX270 FT (JEOL, Tokyo, Japan) and an Avance III HD 600 (Bruker, Yokohama, Japan) as shown in Fig. S2, and the average degree of polymerization (DP) of POEM was calculated by the molar ratio of POEM introduced/Cl moiety.

H (CDCl3): 7.3-6.9 (Ar), 4.08 (COOCH2CH2), 3.8-3.5 (COOCH3 and CH2CH2O), 3.38 (CH2OCH3), 2.1-0.6 ppm (protons of polymer backbone). As a reference material, a block copolymer of MMA and POEM (PMMA-b-PPOEM) was prepared by ATRP, as shown in Fig. 1. EBI (0.2 mmol), CuCl (0.24 mmol), and bpy (0.48 mmol) were added to 50 mL NMP, and the resulting suspension was stirred for 30 min to form a metal-ligand complex for ATRP at ambient temperature. Then, MMA (100 mmol) was added, and molecular oxygen in the reaction mixture was removed by N2 bubbling for 30 min at ambient temperature. Polymerization of the first PMMA block was carried out at 90°C for 24 h, and the product was recovered by precipitation in diethyl ether. The precipitate was dried under vacuum and dissolved in NMP (7 g/100 mL-NMP). CuCl (0.24 mmol), bpy (0.48 mmol) and POEM (28 mmol) were added to the solution and propagation of the second PPOEM block was conducted in the same manner. Finally, the block copolymer was obtained after reprecipitation in diethyl ether. Formation of the block copolymer with 59.0 wt% POEM fraction was confirmed by 1H NMR (Fig. S3), and the yield was 73 %.

H (CDCl3): 4.08 (COOCH2CH2), 3.8-3.5 (COOCH3 and CH2CH2O), 3.38 (CH2OCH3), 2.1-0.6 (methylene protons of polymer backbone), 1.9-1.8 ppm (CCH3).

6

3. TFC Membrane formulation A TFC membrane for gas permeation studies was prepared by spraying a polymeric blend aqueous solution onto a PDMS support. In brief, a 0.4 g hyper-branched polymer was added to 6 mL acetone and dissolved under ultrasonic irradiation in a water bath at 100 W and 37 Hz for 10 min. (Sharp UT106, Osaka, Japan). A 6 mL distilled water was then added to the solution. The resulting solution was mixed with aqueous PVA (7 wt%, 5 mL) to give a polymeric blend solution, where weight fraction of the hyper-branched polymers was 53 wt%. PVA shows excellent membrane formability and has previously been used as a membrane matrix to capture CO2 [34-39], while the polymer does not have CO2-philic nature. The obtained solution was filtered with a membrane filter (nominal pore size: 0.2 µm) and sprayed several times with an air gun-spray on a commercial PDMS sheet to form a CO2-selective layer. The composite membranes were then dried under vacuum at ambient temperature as shown in Fig. S4, and the thickness of the polymeric blend was 8-59 m measured on a Mitutoyo digimatic micrometer (Tokyo, Japan). The polymeric blend for other experiments was prepared by casting the aqueous solution onto a glass Petri dish followed by the same drying process under vacuum at ambient temperature.

4. Gas permeation test Gas permeabilities of the composite membranes were determined by a vacuum time-lag method on a Tsukuba Rika Seiki K-315N-02 (Tokyo, Japan) at 1 atm and 35°C as shown in Fig. S5, where the effective membrane area A was 18.86 cm2, the volume of reservoirs V0 (upper) and V1 (lower) were 22.87 and 22.93 cm3, respectively. After evacuating the membranes in a cell at 50°C for at least 2 h to remove residual gaseous species, the target gas was charged in the upper reservoir at 1 atm and 35°C, and pressure increase dp2/dt in the lower reservoir was then monitored as a function of time. The gas permeability P is calculated from the following equation (Eq. 1). Here, the thickness of composite membranes are l cm, and the dp2(blank)/dt was obtained from a leak test prior to gas permeation experiments. A linear relationship between p2 and t can be found at a steady state, and the slope corresponds to the gas permeability coefficient P. Gas permeance Q is defined as P divided by l (Eq. 2).

P

l V0 273 A 273 T1

V1 1 dp2 273 T2 76 p1 dt

dp2 (blank) dt

(1)

7

P l

Q

(2)

The ideal separation factor for gas A over gas B A/B is the ratio of those gas permeabilities (or permeances), as shown in Eq. 3. A/B

PA PB

QA QB

(3)

5. Membrane characterization Thermal properties of the polymeric materials were examined by DSC on a Shimadzu DSC-60 (Kyoto, Japan). The samples were first heated at 150°C for 10 min. and then cooled down to -100°C with a cooling rate of -20°C/min. to quench the thermal history, and DSC spectra of the second heating were recorded with a heating rate of 10 °C/min from -100 to 150°C. Temperatures of an inflection point are recorded as a glass transition temperature. SAXS experiments were carried out on a Rigaku Nano-Viewer (Rigaku) consisting of a CuK X-ray source (: 1.5418 Å) with a MicroMax-007HF X-ray generator (40 kV and 30 mA), 3 pinhole-collimated beam, and a Pilatus100K 2D detector (Dectris, Baden, Switzerland). The camera length and exposure time were 852 mm and 15 min, respectively. The scattering 2D images were obtained after subtraction of the background scattering spectra by a Fit2D software [40]. The magnitude of the scattering vector q is given by Eq. 4, where the scattering angle was 2. With this experimental set-up, x-ray scattering above 0.156° in scattering angle 2 or 0.111 in scattering vector q can be detected.

q

4

sin

(4)

Results and discussion

1. Preparation of hyper-branched polymers The detail experimental conditions and the results of POEM tethering are described in Table 1.

8

Formation of a hyper-branched structure was confirmed by 1H NMR in Fig. S2. Polymerization of POEM was initiated from the Cl moiety on the PMMA backbone. The triplet resonance of methylene protons adjacent to the ester linkage at 4.30 ppm of monomeric POEM shifted to higher magnetic field at 4.08 ppm, and the peak became broader, which also indicated polymerization of POEM to form a dense POE comb-like structure. Fig. 2 shows the effect of POEM grafting on the macroinitiator backbone. A higher feed ratio of POEM resulted in higher incorporation in the ATRP despite of the Cl moiety introduction between 0.88 to 1.42 mol%. By varying the feed molar ratio of the POEM/Cl moiety from 34.3 to 88.9, the POEM introduction increased from 24.0 to 69.9 wt% in the resulting polymer, as determined by 1H NMR. The obtained results are similar to those reported in previous research in which POEM had a shorter POE chain (Mw 300) [27]. However, the introduction efficiency POEMintroduced/POEMfeed is lower here than that previously reported, due to steric hindrance of the longer POE chain of the monomer (Mw 500), while a longer POE chain would result in producing higher free fractional volume and provide higher gas permeability. When POEM introduction was determined, the average DP of the monomer was calculated by the molar ratio of POEM introduced/Cl moiety, and was 13 to 76. ATRP is known as a living polymerization technique and gives narrow DP distributions, and thus hyper-branched polymers with well-defined chemical structures were prepared with a controlled graft chain distance along the polymer backbone and a controlled graft chain length.



9

Fig. 2. Relation between the feed ratio of POEM in the feed/Cl moiety, and the POEM weight fraction introduced into the hyper-branched polymer. Cl-St in polymer: 0.88 (circle), 1.03 (square), 1.29, (triangle), and 1.42 mol% (diamond).

2. CO2 separation properties of hyper-branched polymers The comb-like structure of the dense POE of hyper-branched polymers is expected to have CO2-philic POE domains with flexible POE chains and has potential in CO2 separation from N2 [27]. A self-standing membrane was readily prepared by solution casting of the polymers, with thicknesses greater than 100 m. The membrane exhibited CO2 permeability of 0.51 GPU (1 GPU=10-6 cm3(STP)/(cm2·s·cmHg)) or 101 barrer, and CO2 selectivity over N2 of 38, when the thickness was 200 m. However, this CO2 permeability should be higher for practical use, which can be achieved by reducing membrane thickness. However, because it is difficult to form free-standing films with a thickness below 80 m, thinner membrane fabrication was carried out on a PDMS support. PDMS has very high gas permeability, with CO2 and N2 permeabilities of 3,490 and 350 barrer, respectively [41,42]. Therefore this support material will not result in major resistance in gas

10

permeation of the hyper-branched polymers. Here, the permeability coefficient of the selective layer was derived by using a series resistance model as follows (Eq. 5) [43,44].

lcomposite Pcomposite

lPDMS PPDMS

lselective Pselective

(5)

where Pcomposite, PPDMS, and Pselective, are permeability coefficients of the composite membranes, a the PDMS support, and the CO2-selective layer, respectively. Accordingly, lcomposite, lPDMS, and lselective are the thicknesses of the composite membrane, PDMS, and selective layer, respectively. The experimentally determined PPDMS was 2,870 barrer for CO2 PPDMS(CO2), and 284 barrer for N2 PPDMS(N2), with CO2/N2 selectivity of 10.1 under the operation conditions. These values are actually lower than the reported values, but the gas permeability of the PDMS support is still much higher than the hyper-branched polymer as a CO2-selective layer, and the selective layer should be a key to determine the CO2 separation performance of the TFC membranes. In the formation of thin CO2-selective layers, the hyper-branched polymer was mixed with PVA to improve the membrane formability and compatibility with the PDMS support. By repeated spraying of the aqueous polymer mixture, the thickness of the selective layers was readily tunable on the support, as listed in Tables 2 and S2. Without the PVA addition, pinhole defects in the TFC membranes were often found in the gas permeation experiment, even when the thickness of the supported selective layer was greater than 100 µm. PVA itself is a gas barrier by nature [45-47], and the gas permeabilities of pure PVA membranes could not be determined within the detection limit of the GC system used. Thus the hyper-branched polymers allowed for preferential CO2 permeation within the selective layer. The CO2 separation performance of the resulting TFC membranes with various CO2-selective layers was measured by a vacuum time-lag method (Tables 2 and S2).

11

Fig. 3. Effect of thickness of composite membranes (a) and CO2-selective layer (b) on the CO2 separation properties (POEM fraction: 24.0 wt%). Open circle: P(CO2), open diamond: P(N2), filled circle: (CO2/N2)

Fig. 3 shows the CO2 separation properties of the TFC membranes and the CO2-selective layer, respectively, as a function of thickness of the selective layer, where POEM fraction in the hyper-branched polymer is 24.0 wt%. The gas permeability coefficients of the TLC membrane Pcomposite appeared to slightly increase with decreasing thickness, and the highest Pcomposite(CO2) was 340 barrer with composite of 20. The average gas permeability coefficients of the CO2 selective layer, Pselective(CO2) and Pselective(N2), were 57.0±10.9 and 1.7±0.8 barrer, respectively, and most likely independent of thickness of the selective layer, as shown in Fig. 3a. The average selectivity of the selective layer was therefore 41.2±19.2. The Qselective(CO2) increased from 2.1 to 5.2 GPU by reducing the thickness from 27 to 12 µm (Table 2). In comparison to the thicker self-standing hyper-branched polymeric membranes, the CO2 permeability is successfully increased by reducing the membrane thickness, while maintaining the CO2 selectivity.

12

Fig. 4. Effect of thickness of composite membranes (a) and CO2-selective layer (b) on the CO2 separation properties (POEM fraction: 40.5 wt%). Open circle: P(CO2), open diamond: P(N2), filled circle: (CO2/N2)

The CO2 permeability is expected to be further improved when a higher amount of POEM is introduced. As shown in Fig. 4, both the gas permeability coefficients of the TLC membranes and the CO2-selective layer were one-order of magnitude higher with 40.5 wt% of POEM fraction, than with 24.0 wt% of POEM fraction. This could be explained by enhanced CO2 solubility with increased POE incorporation. The Pcomposite(CO2) exceeded 1,000 barrer when the thickness of the TFC membranes was below 86 µm, and reached 1,700 barer for a 70 µm membrane in lcomposite (corresponding to 8 µm in lselective). However, both Pcomposite(N2) and Pselective(N2) were also increased by decreasing the corresponding membrane thickness. As a consequence, the selectivity dropped from 80 to 20 in both cases, possibly due to a pinhole defect, often seen when decreasing membrane thickness [48]. However, even in the smallest lselective (8 µm), the CO2-selective layer still maintained moderate CO2 selectivity, and selective was 23.9 with 51.2 GPU in Qselective(CO2), as shown in Table 2. Thus, the CO2 permeance of the hyper-branched polymer was enhanced by a factor of 100 by reducing the membrane thickness from 200 to 8 µm, while the CO2 selectivity decreased from 38 to 24.

13

Fig. 5. Effect of thickness of composite membranes (a) and CO2-selective layer (b) on the CO2 separation properties (POEM fraction: 69.9 wt%). Open circle: P(CO2), open diamond: P(N2), filled circle: (CO2/N2)

Figs. 5 shows the CO2 separation properties of TFC membranes and the selective layer, respectively, where the weight fraction of POEM is 69.9 wt% in the hyper-branched polymers. The Pcomposite(CO2) was over 1,000 barrer when the lcomposite was 74 µm (12 µm lselective). The Pselective(CO2) was also higher than those with lower POEM fraction (24.0 wt%), and the CO2 selectivity was larger than 30. Similar to the above results, both of the Pcomposite and the Pselective increased as the corresponding membrane thickness decreased. Since the change in permeability coefficient as a function of membrane thickness –dP/dl of N2 is grater than that of CO2, the CO2 selectivity decreased by reducing the thicknesses of the corresponding membranes. Compared with the thicker self-standing membranes (thickness: 200 µm), the hyper-branched polymers on a PDMS support exhibited higher CO2 permeability with relatively low CO2 selectivity, which may result from formation of pinhole defects with a thinner CO2-selective layer. However, the defects are not serious, and the selective layer still maintains moderate CO2 selectivity of >20, similar to various polyimides developed for CO2 capture [49], including polymers of intrinsic microscopy PIM-1 [50]. It is clear that higher POEM fraction >40 wt% resulted in higher CO2 permeability, and that the Pcomposite exceeded 1,000 barrer for a lselective smaller than 24 µm. This is expected to be due to the formation of a POE-rich domain upon microphase separation between POE and other polymeric factions, resulting in an increase in the CO2 solubility in the selective layer,

14

corresponding to higher CO2 permeability. On the other hand, molecular architecture by self-assembly is also an important factor for the separation properties of polymeric membranes. The hyper-branched polymers displayed higher CO2 permeability (ca. 1,400 barrer) than block copolymers (850 barrer) with similar membrane thickness and POEM fraction, while the selectivity was slightly lower, as shown in Table 2. Tsang, et al. compared the proton conductivity of block copolymers of poly(vinylidene difluoride) (PVDF) and sulphonated polystyrene (SPS), and showed that PVDF-g-SPS had higher proton conductivity than PVDF-b-SPS due to higher chain mobility of the SPS block [51]. The difference in the molecular architecture was also simulated for an amphiphilic block copolymer for fuel cell membranes by Dorenbos and Morohoshi [52]. At equal ion exchange capacity of a hydrophilic polymeric block, a graft polymer displays a higher water diffusion constant. This implies that the hydrophilic graft chain is more flexible than that in the linear block copolymers. The tethered dense POEs comb onto the PMMA backbone would have higher chain mobility and thus result in giving higher CO2 diffusivity than that within the polymer backbone.

3. Membrane characterizations Although various POE-containing polymers have been used for matrices of CO2 separation membranes, high molecular weight POE (Mn >2,000) shows more crystalline nature, which often reduces the CO2 separation performance [20,53]. The average EO units of POEM used in this study was 9, and thus the POE is not considered crystalline, and flows under ambient conditions. Meanwhile, the PMMA backbone was in a glassy state, affording mechanical strength to the hyper-branched polymers under the operation conditions of the gas permeation experiments. On the other hand, low compatibility between POE and PMMA is expected to induce phase separation on a submicron scale in the polymeric membranes, and thus CO2 would preferentially diffuse through the resulting POE-rich domain upon microphase separation in the membranes. Such microphase separation was investigated by DSC and SAXS. Fig. 6 displays typical DSC thermograms of PMMA-g-PPOEM hyper-branched polymers (POEM weight fractions: 31.0 and 60.1 wt%) and PMMA-b-PPOEM linear block copolymer

15

(POEM weight fractions: 59.9 wt%). For the hyper-branched polymers, a thermal transition was seen at -58°C when the POEM fraction was 60.1 wt% (bold line), although no transitions were found for the specimen with 31.0 wt% of POEM (plain line). The hyper-branched polymer with higher POE fraction (60.1 wt%) also displayed a thermal transition, when it was mixed with PVA (extrabold line). On the other hand, PMMA-b-PPOEM displayed a thermal transition at -59°C (dashed line), which was very close to those seen in the polymeric samples with 60.1 wt% of POEM and corresponded to the glass transition of POE. The obtained result strongly suggests the presence of a POE-rich domain upon phase separation between POE and the other polymeric fractions for the higher POEM introduced polymeric specimens. On the contrary, when the POEM fraction is 31.0 wt%, the DSC result indicated that POE was miscible with the other polymers, and the polymer no longer underwent the glass transition of POE. The repulsive force of the lower POE weight fraction toward PMMA and other fractions was not high enough to form a POE-rich domain by the phase separation.

Fig. 6. DSC thermograms of various POE-containing polymers with a heating rate of 10°C/min. Plain: hyper-branched polymers with 31.0 wt% of POEM, bold: hyper-branched polymers with 60.1 wt% of POEM, extrabold: blend of hyper-branched polymers with 60.1 wt% of POEM and PVA,

16

dashed: PMMA-b-PPOEM block copolymer with 59.9 wt% of POEM

The hyper-branched polymer was blended with PVA for the gas separation test to form a CO2 selective layer on a PDMS support, as described above. The blend with higher POEM fraction (60.1 wt%) also underwent a glass transition corresponding to POE at -57°C, which was identical to the non-blended specimen. The result supports the formation of a POE-rich domain in the selective layer, and that results in enhancement of the CO2 permeability. DSC is often used for validation of microphase separation of block copolymers. However, spatial information of the phase separation, such as a periodic structure, is not generally provided. In addition, heating might induce phase separation. SAXS can detect the difference in electron density on a submicron scale, which corresponds well to the structural length of microphase separation of BCPs. Thus, microphase separation of hyper-branched polymers was studied by SASX, as seen in Fig. 7. The average structural length  was determined from a peak at qmax by fast Fourier transformation of a scattering pattern of samples, as calculated from Eq. 6.

2 q

(6)

Fig. 7. SAXS spectra of PMMA-g-PPOEM hyper-branched polymers with 31.0 (a) and 60.1 wt% of POEM (b), PMMA-b-PPOEM block copolymer with 59.9 wt% of POEM (c), and blend of PMMA-g-PPOEM with 60.1 wt% of POEM and PVA (d)

17

When the POEM weight fraction was ca. 60 wt%, scattering peaks were observed at 0.193 (PMMA-g-PPOEM) and 0.280 nm-1 (PMMA-b-PPOEM) in q (Figs 7b and c), which indicated that those polymers formed periodic structures upon microphase separation between POE and the other polymeric fractions. The  values of the hyper-branched polymer (POEM: 60.1 wt%) and block copolymer (POEM: 59.9 wt%) were calculated to be 32.5 and 22.4 nm, respectively, using Eq. 6. This result clearly indicates the presence of POE-rich domain upon the microphase separation. The graft copolymer represents a larger structural length than the block copolymer with similar POE weight fraction, although the detail has not been elucidated. A flexible, dense POE comb graft would be expected to provide a greater periodic length than the confined POE comb on the polymer backbone, as discussed in the above section, which would result in the higher CO2 transportation properties seen in Table 2. The effect of PVA addition on the microphase separation of PMMA-g-PPOEM was also examined, as shown in Fig. 7d. The polymeric blend also displayed a peak in the SAXS experiment, and a shift in  was observed from 32.5 to 47.0 nm. However, PVA does not interfere the formation of a POE-rich domain (and thus CO2 permeation) as confirmed by DSC in Fig. 6, while PVA is not related to gas permeation at all. On the other hand, the hyper-branched polymer with lower POEM fraction (POEM: 31.0 wt%, Fig. 7a) did not result in peaks derived from periodic structures. The lower POEM fraction was not sufficient to induce phase separation. The obtained results corresponded well with DSC, and higher POE incorporation resulted in the formation of a POE-rich domain. The gas permeation experiment suggested that higher POE weight fraction displayed higher CO2 transport properties. Thus, formation of the CO2-philic POE-rich domain in the CO2-selective layer is crucial to enhance the gas permeability.

Conclusion

Hyper-branched polymers with longer POE chain (average oxyethylene unit: 9) were

18

successfully prepared by ATRP of the corresponding POEM from Cl moiety of PMMA backbone macroinitiator. The POEM weight fraction ranged from 24.0 to 69.9 wt%, where the average DP of POEM was 6.1 to 45.6. A thin film layer of the polymer was formed on a high gas permeable PDMS support with the formation of a CO2 selective layer, and the thickness was readily tunable (8-59 µm). Blending of PVA with the polymer can suppressed the formation of major pinhole defects in the selective layer preparation, and the resulting TFC membranes showed much higher CO2 transport properties than the thicker self-standing membranes of the hyper-branched polymers with moderate CO2 selectivity, comparable to various polyimides developed for CO2 separation. The CO2 permeability coefficients of the composite membranes were increased to greater than 1,000 barrer by reducing the thickness of the selective layer to less than 25 µm. With 40.5 wt% of POE fraction, the permeability coefficient reached 1,470 barrer with CO2 selectivity of 24.5, when the thickness of the selective layer was 15 µm. However, the CO2 permeability coefficient was not high with a lower POEM content (24.0 wt%). This could be explained by the difference in CO2 solubility in the selective layers. When the POEM fraction is below 31.0 wt%, the POE is miscible in the polymer matrix, as investigated by DSC and SAXS. The CO2 solubility of the POE-PMMA mixture was lower than that of a POE-rich domain, where CO2 can pass through preferentially with higher permeability. A dense POE comb formation is promising to fabricate CO2 separation membranes. Grajales et al. reported the fabrication of PPOEM on a porous alumina and the CO2 separation performance over H2 [54]. The CO2 permeability coefficient ranged from 20 to 60 barrer with CO2/H2 selectivity of 12, when the CO2 selective layer was less than 500 nm in thickness. In comparison to organic/inorganic hybrid membranes, polymeric membranes could be suitable for CO2 capture at a mass emission sources in terms of large-scale production, homogeneity, processability, and cost. The hyper-branched polymers actually exhibited much higher CO2 permeability than hybrid membranes, despite the larger thickness. Furthermore, it was revealed that the dense POE comb structure on the graft chain of the hyper-branched architecture resulted in higher CO2 permeation than that integrated into the polymer backbone. The dense comb graft structure gives greater chain flexibility in the POE-rich domain upon microphase separation, where CO2 migrates with a higher diffusivity.

19

Acknowledgements The authors acknowledge Prof. Anne M. Mayes of Massachusetts Institute of Technology for invaluable support to develop the hyper-branched polymers. A part of this research is supported by the Japan Science and Technology Agency-Advanced Low Carbon Technology Research and Development Program (JST-ALCA). The authors also thank Prof. Takashi Aoki of Kyoto Institute of Technology for helping SAXS measurements, Profs. Yu Hoshino and Yoshiko Miura of Kyushu University for DSC measurements, and Prof. Stephen M. Lyth of Kyushu University for editing the manuscript.

References [1] Nikos Hadjichristidis, Stergios Pispas, George Floudas, Block Copolymers: Synthetic Strategies, Physical Properties, and Applications, Wiley, New Jersey, 2003. [2] S.B. Darling, Directing the self-assembly of block copolymers, Prog. Polym. Sci. 32 (2007) 1152-1204. [3] V. Balasubramanian, B. Herranz-Balanco, P.V. Almeida, J. Hirvonen, H.A. Santos, Multifaceted polymersome platforms: Spanning from self-assembly to drug delivery and protocells, Prog. Polym. Sci. 60 (2016) 51-85. [4] Z. Tanga, C. Hea, H. Tiana, J. Dinga, B.S. Hsiaob, B. Chub, X. Chen, Polymeric nanostructured materials for biomedical applications, Prog. Polym. Sci. 60 (2016) 86-128. [5] C. Park, J. Yoon, E.L. Thomas, Enabling nanotechnology with self assembled block copolymer

patterns, Polymer 44 (2003) 6725-6760. [6] C. Cummins, T. Ghoshal, J.D. Holmes, M.A. Morris, Strategies for Inorganic Incorporation using Neat Block Copolymer Thin Films for Etch Mask Function and Nanotechnological Application, Adv. Mater. 28 (2016) 5586-5618. [7] S.P. Nunes, A. Car, From Charge-Mosaic to Micelle Self-Assembly: Block Copolymer Membranes in the Last 40 Years, Ind. Eng. Chem. Res. 52 (2013) 993-1003.

20

[8] V.I. Bondar, B.D. Freeman, I. Pinnau, Gas sorption and characterization of poly(ether-b-amide) segmented block copolymers, J. Polym. Sci. Part B: Polym. Phys. 37 (1999) 2463-2475. [9] V.I. Bondar, B.D. Freeman, I. Pinnau, Gas transport properties of poly(ether-b-amide) segmented block copolymers, J. Polym. Sci. Part B: Polym. Phys. 38 (2000) 2051-2062. [10] J.H. Kim, S.Y. Ha, Y.M. Lee. Gas permeation of poly(amide-6-b-ethylene oxide) copolymer, J. Membr. Sci. 190 (2001) 179-193. [11] H. Lin, B.D. Freeman, Materials selection guidelines for membranes that remove CO2 from gas mixtures, J. Mol. Str. 739 (2005) 57-74. [12] S.J. Metz, M.H.V. Mulder, M. Wessling, Gas-permeation properties of poly(ethylene-oxide) poly(butylene terephthalate) block copolymers, Macromolecules 37 (2004) 4590-4597. [13] A. Car, C. Stropnik, W. Yave, K.V. Peinemann, Tailor-made polymeric membranes based on segmented block copolymers for CO2 separation, Adv. Funct. Mater. 18 (2008) 2815-2823. [14] W. Yave, A. Szymczyk, N. Yave, Z. Roslaniec, Design, synthesis, characterization and optimization of PTT-b-PEO copolymers: a new membrane material for CO2 separation, J. Membr. Sci. 362 (2010) 407-416. [15] K. Okamoto, M. Fujii, S. Okamyo, H. Suzuki, K. Tanaka, H. Kita, Gas permeation properties of poly(ether imide) segmented copolymers, Macromolecules 28 (1995) 6950-6956. [16] H.Z. Chen, Y.C. Xiao, T.S. Chung, Synthesis and characterization of poly(ethylene oxide) containing copolyimides for hydrogen purification, Polymer 51 (2010) 4077-4086. [17] A. Tena, A. Marcos-Fernandez, A.E. Lozano, J.G. de la Campa, J. de Abajo, L. Palacio, P. Pradanos, A. Hernandez, Thermally treated copoly(ether-imide)s made from bpda and alifatic plus aromatic diamines. Gas separation properties with different aromatic diamimes, J. Membr. Sci. 387 (2012) 54-65. [18] L.S. Teo, J.F. Kuo, C.Y. Chen, Study on the morphology and permeation property of amine group-contained polyurethanes, Polymer 39 (1998) 3355-3364. [19] H.W. Kim, H.B. Park, Gas diffusivity solubility and permeability in polysulfone-poly(ethylene oxide) random copolymer membranes, J. Membr. Sci. 372 (2011) 116-124. [20] S.L. Liu, L. Shao, M.L. Chua, C.H. Lau, H. Wang, S. Quan, Recent progress in the design of advanced PEO-containing membranes for CO2 removal, Prog. Polym. Sci. 38 (2013) 1089-1120. [21] H. Lin, B.D. Freeman, Gas solubility, diffusivity and permeability in poly(ethylene oxide), J. Membr. Sci. 239 (2004) 105-117. [22] H. Lin, T. Kai, B.D. Freeman, S. Kalakkunnath, D.S. Kalika,The Effect of cross-linking on gas permeability in cross-linked poly(ethylene glycol diacrylate), Macromolecules 38 (2005) 8381-8393.

21

[23] H. Lin, B.D. Freeman, Gas and vapor solubility in cross-linked poly(ethylene glycol diacrylate), Macromolecules 38 (2005) 8394-8407. [24] H. Lin, E.V. Wagner, B.D. Freeman, L.G. Toy, R.P. Gupta, Plasticization-enhanced hydrogen purification using polymeric membranes, Science 311 (2006) 639-642. [25] H.Q. Lin, B.D. Freeman, Gas permeation and diffusion in crosslinked poly(ethylene glycol diacrylate), Macromolecules 39 (2006) 3568-3580. [26] H. Lin, E.V. Wagner, J.S. Swinnea, B.D. Freeman, S.J. Pas, A.J. Hill, S. Kalakkunnath, D.S. Kalika, Transport and structural characteristics of crosslinked poly(ethylene oxide) rubbers, J. Membr. Sci. 276 (2006) 145-161. [27] I. Taniguchi, K. Kinugasa, S. Egashira, M. Higa, Preparation of well-defined hyper-branched polymers and the CO2 separation performance, J. Membr. Sci. 502 (2016) 124-132. [28] J.H. Koh, K.J. Lee, J.A. Seo, J.H. Kim, Amphiphilic polymer electrolytes consisting of PVC-g-POEM comb-like copolymer and LiCF3SO3, J. Polym. Sci. Part B: Poym. Phys. 47 (2009) 1443-1451. [29] J.D. Figueroa, T.Fout, S. Plasynski, H. McIlvried, R.D. Srivastava, Advances in CO2 capture technology-The U.S. Department of Energy’s Carbon Sequestration Program, Int. J. Greenh. Gas Control 2 (2008) 9-20. [30] T.C. Merkel, H. Lin, X. Wei, R. Baker, Power plant post-combustion carbon dioxide capture: An opportunity for membranes, J. Membr. Sci. 359 (2010) 126-139. [31] W.A. Braunecker, K. Matyjaszewski, Controlled/living radical polymerization: Features, developments, and perspectives, Prog. Polym. Sci. 32 (2007) 93-146. [32] Z. Zhang, P. Zhang, Y. Wang, W. Zhang, Recent advances in organic–inorganic welldefined hybrid polymers using controlled living radical polymerization techniques, Polym. Chem. 7 (2016) 3950-3976. [33] M. Ding, X. Jiang, L. Zhang, Z. Cheng, X. Zhu, Recent Progress on Transition Metal Catalyst Separation and Recycling in ATRP, Macromol. Rapid Commun. 36 (2015) 1702-1721. [34] H. Matsuyama, A. Terada, T. Nakagawara, Y. Kitamura, M. Teramoto, Facilitated transport of CO2 through polyethylenimine/poly(vinyl alcohol) blend membrane, J. Membr. Sci. 163 (1999) 221-227. [35] J. Huang, J. Zou, W.S.W. Ho, Carbon Dioxide Capture Using a CO2-Selective Facilitated Transport Membrane, Ind. Eng. Chem. Res. 47 (2008) 1261-1267. [36] L. Deng, T.-J. Kim, M.-B. Hägg, Facilitated transport of CO2 in novel PVAm/PVA blend membrane, J. Membr. Sci. 340 (2009) 154-163. [37] R. Xinga, W.S.W. Ho, Crosslinked polyvinylalcohol–polysiloxane/fumed silica mixed matrix membranes containing amines for CO2/H2 separation, J. Membr. Sci. 367 (2011) 91-102.

22

[38] S. Duan, I. Taniguchi, T. Kai, S. Kazama, Poly(amidoamine) dendrimer/poly(vinyl alcohol) hybrid membranes for CO2 capture, J. Membr. Sci. 423-424 (2012) 107-112. [39] V. Vakharia, K. Ramasubramanian W.S.W. Ho, An experimental and modeling study of CO2-selective membranes for IGCC syngas purification, J. Membr. Sci. 488 (2015) 56-66. [40] A freeware computer program developed by A. Hammersley,

http://www.esrf.eu/computing/scientific/ FIT2D/. [41] S.A. Stern, V.M. Shah, B.J. Hardy, Structure-permeability relationships in silicone polymers, J. Polym. Sci. Part B Polym. Phys. 25 (1987) 1263-1298. [42] S.A. Stern, B.D. Bhide, Permeability of Silicone Polymers to Ammonia and Hydrogen Sulfide, J. Appl. Polym. Sci. 38 (1989) 2131-2147. [43] D.R. Paul, Gas transport in homogeneous multicomponent polymers, J. Membr. Sci. 18 (1984) 75-86. [44] B.W. Rowe, B.D. Freeman, D.R. Paul, Physical aging of ultrathin glassy polymer films tracked by gas permeability, Polymer 50 (2009) 5565-5575. [45] J.-H Yeun, G.-S. Bang, B.J. Park, S.K. Ham, J.-H. Chang, Poly(vinyl alcohol) Nanocomposite Films: Thermooptical Properties, Morphology, and Gas Permeability, J. App. Polym. Sci. 101 (2006) 591-596. [46] H.M. Kim, J.K. Lee, H.S. Lee, Transparent and high gas barrier films based on poly(vinyl alcohol)/graphene oxide composites, Thin Solid Films 519 (2011) 7766-7771. [47] J. Gaume, P. Wong-Wah-Chung, A. Rivaton, S. Therias, J.-L. Gardette, Photochemical behavior of PVA as an oxygen-barrier polymer for solar cell encapsulation, RSC advances 1 (2011) 1471-1481. [48] R.R. Tiwari, Z.P. Smith, H. Lin, B.D. Freeman, D.R. Paul, Gas permeation in thin films of

“high free-volume” glassy perfluoropolymers: Part I. Physical aging, Polymer 55 (2014) 5788-5800. [49] S. Wang, X. Li, H. Wu, Z. Tian, Q. Xin, G. He, D. Peng, S. Chen, Y. Yin, Z. Jiang, M.D.

Guiver, Advances in high permeability polymer-based membrane materials for CO2 separations, Energy Environ. Sci. 9 (2016) 1863-1890. [50] P.M. Budda, K.J. Msayib, C.E. Tattershall, B.S. Ghanem, K.J. Reynolds, N.B. McKeown, D. Fritsch, Gas separation membranes from polymers of intrinsic microporosity, J. Membr. Sci. 251 (2005) 263-269. [51] E.M.W. Tsang, Z. Zhang, Z. Shi, T. Soboleva, S. Holdcroft, Considerations of Macromolecular Structure in the Design of Proton Conducting Polymer Membranes: Graft versus Diblock Polyelectrolytes, J. Am. Chem. Soc. 129 (2007) 15106-15107.

23

[52] G. Dorenbos, K. Morohoshi, Chain architecture dependence of pore morphologies and water diffusion in grafted and block polymer electrolyte fuel cell membranes, Energy Environ. Sci., 3 (2010) 1326-1338. [53] K. Tsutsui, H. Yoshimizu, Y. Tsujita, T. Kinoshita, Gas permeation properties of a composite membrane filled with poly(ethylene oxide) into a porous membrane, J. App. Polym. Sci. 73 (1999) 2733-2738. [54] S.T. Grajales, X. Dong, Y. Zheng, G.L. Baker, M.L. Bruening, Effects of Monomer Composition on CO2-Selective Polymer Brush Membranes, Chem. Mater. 22 (2010) 4026-4033.

Table 1. Synthetic conditions of hyper-branched polymers and the results of polymerization Cl-St in Run

polymer /mol%

MI /g

CuCl

bpy

-4

-4

POEM -4

Reaction

/10

/10

/10

time

mol

mol

mol

/h

Yield

DP of

/%

POEM

POEM fraction /wt%

1

0.88

3.01

3.7

7.2

220

23

71.1

14.0

38.0

2

0.88

3.01

3.7

7.2

220

25

54.0

15.5

40.5

3

0.88

3.00

3.3

6.5

220

24

58.6

15.5

40.5

4

0.88

3.00

3.3

6.5

220

24

50.9

18.3

44.6

5

1.03

1.50

1.6

3.0

55

24

46.0

6.1

24.0

6

1.03

1.50

1.6

2.9

110

24

51.6

10.3

34.7

7

1.02

3.00

3.5

6.4

271

48

61.9

45.6

69.9

8

1.42

3.00

3.6

7.3

286

24

52.8

10.4

42.4

9

1.42

3.00

3.6

7.3

286

48

66.5

10.7

43.0

10

1.29

3.00

4.6

9.2

132

48

64.2

7.1

31.5

11

1.29

3.00

5.2

10.1

132

48

78.1

8.1

34.4

12

1.29

3.00

4.6

9.3

220

48

65.2

8.3

34.7

13

1.29

3.00

4.7

9.2

264

48

41.6

7.7

33.1

14

1.29

3.00

5.2

10.1

264

48

74.9

11.1

41.6

15

1.29

3.00

4.6

9.2

286

48

52.2

11.3

42.0

16

1.29

3.00

4.7

9.2

286

49

65.8

12.9

45.3

17

1.29

3.01

5.1

9.3

308

72

56.8

33.5

68.3

24

Table 2. CO2 separation properties of PMMA-g-PPOEM hyper-branched polymer and PMMA-b-PPOEM block copolymer composite membranes at 35°C PO Ru EM n

/wt %

3-1 24.

lcomp

lselec tive

osite

/µm

/µ m

Pcomposite

Pcomposit

(CO2)

e(N2)

/barrer

/barrer

com posite

Pselective(

Pselectiv

Qselective

CO2)

e(N2)

(CO2)

/barrer

/barrer

/GPU

Qselectiv e(N2) 2

/×10

sele ctive

GPU

80

18

282

4.51

62.5

68.7

1.03

3.82

5.71

66.8

89

27

179

2.54

70.5

56.8

0.78

2.10

2.87

73.2

74

12

347

11.7

29.7

62.6

1.97

5.22

16.4

31.9

78

16

290

11.1

26.1

64.7

2.35

4.04

14.7

27.5

76

14

304

13.6

22.4

61.3

2.61

4.38

18.6

23.5

109

47

468

6.67

70.2

222

2.91

4.73

6.20

76.3

85

23

1,000

25.4

39.4

363

7.35

15.8

32.0

49.3

80

18

1,340

54.2

24.7

473

14.3

26.3

79.5

33.0

86

24

1,090

34.6

31.5

419

10.6

17.5

44.1

39.6

75

13

1,380

51.2

27.0

397

10.4

30.5

80.2

38.1

77

15

1,470

60.0

24.5

487

14.1

32.5

93.9

34.6

77

15

1,480

66.8

22.2

492

16.1

32.8

107

30.6

121

59

320

4.01

79.8

166

1.97

2.80

3.34

84.0

93

31

735

14.3

51.4

305

4.93

9.84

15.9

61.9

70

8

1,700

102

16.7

409

17.1

51.2

214

23.9

74

12

1,380

62.0

22.3

374

12.3

31.2

103

30.4

74

12

1,090

35.9

30.4

260

6.51

21.7

54.3

39.9

0 3-2 24. 0 3-3 24. 0 3-4 24. 0 3-5 24. 0 5-1 40. 5 5-2 40. 5 5-3 40. 5 5-4 40. 5 5-5 40. 5 5-6 40. 5 5-7 40. 5 5-8 40. 5 5-9 40. 5 5-1 40. 0

5

7-1 69. 9 7-2 69. 9

25

7-3 69.

100

38

515

9.07

56.8

220

3.52

5.79

9.25

62.6

112

50

415

5.33

77.9

201

2.40

4.03

4.81

83.8

60

14

856

24.8

34.5

280

6.37

19.7

44.9

44.0

9 7-4 69. 9 BC 59. P

9

i=Pi(CO2)/Pi(N2), 1 GPU=10-6 cm3(STP)/(cm2·s·cmHg), 1 barrer=10-10 cm3(STP)cm/(cm2·s·cmHg)

GRAPHICAL ABSTRACT

26

Highlights 

Hyper-branched polymers with dense POE comb graft can be prepared.



The polymers show preferential CO2 permeation over N2.



Thin film composite membranes give high CO2 permeability over 1,000 barrer.



POE-rich domain upon microphase separation is confirmed by DSC and SAXS.



Formation of POE-rich domain is crucial for high CO2 separation performance.

27