Dioxolane-Based Perfluoropolymers with Superior Membrane Gas

6 days ago - Department of Chemical and Biomolecular Engineering, Tandon School of Engineering, New York University, Six MetroTech Center, Brooklyn , ...
0 downloads 11 Views 2MB Size
Download PDF
Article Cite This: Macromolecules XXXX, XXX, XXX−XXX

Dioxolane-Based Perfluoropolymers with Superior Membrane Gas Separation Properties Milad Yavari,† Minfeng Fang,‡ Hien Nguyen,† Timothy C. Merkel,§ Haiqing Lin,*,† and Yoshiyuki Okamoto*,‡ †

Department of Chemical and Biological Engineering, University at Buffalo, The State University of New York, Buffalo, New York 14260, United States ‡ Department of Chemical and Biomolecular Engineering, Tandon School of Engineering, New York University, Six MetroTech Center, Brooklyn, New York 11201, United States § Membrane Technology and Research, Inc., 39630 Eureka Drive, Newark, California 94560, United States S Supporting Information *

ABSTRACT: Glassy perfluoropolymers have become an exciting materials platform for membrane gas separation as they define the upper bounds for some gas separations, such as He/H2, He/CH4, and N2/CH4. However, due to the difficulty in synthesis, only a few glassy perfluoropolymers are commercially available, including Teflon AF and Hyflon AD derived from dioxoles and Cytop derived from dihydrofuran. In this study, two perfluoropolymers based on dioxolanes, poly(perfluoro-2-methylene-1,3-dioxolane) (poly(PFMD)) and poly(perfluoro-2-methylene-4-methyl-1,3-dioxolane) (poly(PFMMD)), were synthesized by radical polymerization and characterized thoroughly for physical properties such as glass transition temperature (Tg), d-spacing between polymer chains, and fractional free volume (FFV). The gas permeability and solubility were determined at 35 °C for a series of pure gases in these perfluorodioxolanes and compared with the commercial perfluoropolymers. Poly(PFMD) and poly(PFMMD) exhibit separation properties of He/H2, He/CH4, H2/CH4, H2/CO2, and N2/CH4 near or above the upper bounds in Robeson’s plots, and superior to the commercial perfluoropolymers, despite their similar Tg and FFV. The underlying reasons for the superior gas separation properties in these dioxolane-based perfluoropolymers are discussed.

■

tions.5,11,17,18 These polymers were successfully fabricated into defect-free TFC membranes and showed superior separation properties for He/gas and H2/CH4. For example, TFC membranes based on poly(perfluoro-2-methylene-4methyl-1,3-dioxolane) (poly(PFMMD) with the chemical structure shown in Figure 1) showed a He permeance of 2160 GPU (1 GPU = 10−6 cm3 (STP)/(cm2 s cmHg)) and He/CH4 selectivity of 108, which represent a significant improvement over the properties of membranes based on Hyflon AD60 or Cytop for He recovery from natural gas.5 Because the thickness of the selective layer in the TFC membranes cannot be accurately determined, the intrinsic gas permeability in these perfluoropolymers cannot be calculated. As a result, there is a lack of systematic fundamental gas transport properties such as gas permeability, solubility, and diffusivity for these dioxolane-based perfluoropolymers, which could be used for comparison with commercially available

INTRODUCTION Amorphous glassy perfluoropolymers have attracted significant interest for membrane gas separation due to their superior separation properties. Currently, the most widely studied ones are commercially available Teflon AF, Hyflon AD, and Cytop with chemical structures shown in Figure 1.1−7 These perfluoropolymers have high fractional free volume (FFV), leading to high gas permeability, and unfavorable interactions with H2 and hydrocarbons, resulting in high He/gas and gas/ CH4 selectivities. Consequently, these perfluoropolymers determine the upper bounds for a variety of gas separations in Robeson’s plots, such as He/H2, He/CH4, He/N2, He/CO2, and H2/CH4.8−11 Moreover, the amorphous perfluoropolymers can be easily fabricated into industrial thin film composite (TFC) membranes (as thin as 50 nm) with good resistance to physical aging; i.e., gas permeability remains relatively stable over time.12−16 However, there are only a few amorphous perfluoropolymers available for membrane applications, mainly due to the difficulty in the synthesis of these polymers. Recently, we have reported a new series of fluoropolymers based on dioxolanes for use in membrane gas separa© XXXX American Chemical Society

Received: February 4, 2018 Revised: March 7, 2018

A

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 2. Polymerization of perfluorodioxolanes (R/R′ = CF3/F for PFMMD or F/F for PFMD). immiscible with the perfluorinated monomers or solvents and they form cages, which retard chain initiation.20 To determine the average molecular weight (Mv) of poly(PFMMD), the intrinsic viscosity ([η]) of the polymer solution was derived using the Huggins equation, and the viscosities of a series of dilute polymer solutions were measured in a Ubbelohde viscometer.21 Film Preparation. Poly(PFMMD) films with a thickness of about 50 μm were prepared by a solution casting method using 5 wt % poly(PFMMD) in HFB. After drying in air, the films were further dried at 80 °C under vacuum overnight to remove residual HFB. Poly(PFMD) films with a thickness of about 30 μm were prepared by a melt-press method because this polymer is not soluble in any solvents. Specifically, polymer powders were sandwiched between Kapton films (McMaster-Carr, Elmhurst, IL) and pressed at 250 °C using a hydraulic press (MTI Corporation, Richmond, CA). Freestanding films of Teflon AF1600 and Hyflon AD80 were prepared using the solution casting method. After drying in air, the films were further dried at 80 °C under vacuum overnight to remove any residual solvent (i.e., Novec 7200). To eliminate the effect of film preparation history on polymer morphology, the dried films were annealed under vacuum at temperatures above their respective Tg, i.e., 160 °C for Hyflon AD80 and 190 °C for Teflon AF1600 for 20 min before a rapid cooling to 20 °C. Cytop films were prepared by solution casting. After drying in air, the films were further dried at 130 °C under vacuum and then sandwiched between Kapton films followed by melt-pressing at 160 °C. The hot pressing is an effective approach to eliminate pinholes in the films. The obtained films were annealed at 130 °C (above their Tg of 108 °C) for 30 min to mitigate the effect of the film processing history on physical properties.22 Characterization of Physical Properties of Polymer Films. The density of polymer films was determined using Archimedes’ principle and an analytical balance (Model XS 64, Mettler-Toledo, Columbus, OH) equipped with a density kit.23 The films were first weighed in air and then in n-decane at 21 °C. The obtained weight values are used to calculate the polymer density, using the density of ndecane (0.73 g/cm3). An Ultima IV X-ray diffractometer (Rigaku Corporation, Tokyo, Japan) with Cu Kα radiation (a wavelength of 1.54 Å) was used to characterize the crystallinity and d-spacing of the polymer films. The scanning ranged from 5° to 50° at 0.5°/min. Differential scanning calorimetry (DSC Q2000, TA Instruments, New Castle, DE) was used to determine the thermal transitions of the polymers at a heating rate of 10 °C/min under a nitrogen flow of 50 mL/min. The Tg and melting temperature (Tm) were determined using Universal Analysis 2000 software. Determination of Pure-Gas Permeability and Sorption Properties. The pure-gas permeability of pristine polymer films to CH4, N2, Ar, CO2, H2, and He was determined using a constant volume/variable pressure apparatus at 35 °C.14,24 The polymer film was masked with aluminum tape and mounted in a permeation cell (Millipore Corporation, Bedford, MA). The system was evacuated overnight, and then gas at the desired pressure was introduced to the upstream volume. The pressure increase in the downstream chamber with a known volume was recorded as a function of time. The steadystate rate of pressure increase in the downstream volume was used to calculate gas permeability (PA). PA has units of barrer, where 1 barrer = 10−10 cm3 (STP) cm/(cm2 s cmHg). The pure-gas solubility of CH4, Ar, CO2, C2H4, and C2H6 in the polymer films was determined by a pressure decay method using a dual-volume and dual-transducer apparatus at 35 °C.24,25 Approx-

Figure 1. Chemical structures of perfluoropolymers including commercial dioxole-derived Teflon AF and Hyflon AD, and dihydrofuran-derived Cytop, and the two dioxolane-based perfluoropolymers synthesized in this study, poly(PFMD) and poly(PFMMD).

perfluoropolymers. The intrinsic properties of these materials are critical to derive polymer structure/property relationships, which can be used to guide the design of next-generation perfluoropolymers for membrane gas separation. In this study, for the first time, we determined gas solubility and permeability of poly(PFMMD) and poly(perfluoro-2methylene-1,3-dioxolane) (poly(PFMD)) at 35 °C and calculated gas diffusivity. Poly(PFMD) is an analogue to poly(PFMMD) without the −CF3 substituent on the dioxolane ring, as shown in Figure 1. The transport properties are correlated with the physical properties of the perfluoropolymers such as glass transition temperature (Tg), fractional free volume (FFV), and d-spacing. The gas separation properties of these dioxolane-based perfluoropolymers are also compared with commercially available perfluoropolymers with similar Tg and FFV, and a possible mechanism for the superior gas separation properties in the dioxolane-based perfluoropolymers is elucidated. The superior gas separation properties in poly(PFMD) and poly(PFMMD) are also demonstrated in Robeson’s upper bound plots.

■

EXPERIMENTAL SECTION

Materials. Teflon AF1600, Hyflon AD80, and Cytop (in solutions) were purchased from DuPont (Wilmington, DE), Solvay Specialty Polymers (West Deptford, NJ), and Asahi Glass Co. (Tokyo, Japan), respectively. Engineered fluid Novec 7200 (ethoxynonafluorobutane) was purchased from 3M (St. Paul, MN) and used to dissolve Teflon AF1600 and Hyflon AD80. Hexafluorobenzene (HFB, 99%) was purchased from Sigma-Aldrich (St. Louis, MO) and used to dissolve poly(PFMMD). Gas cylinders of N2, Ar, H2, and He with 99.999% purity, CH4 and CO2 with 99.9% purity, and C2H6 and C2H4 with 99.5% purity were obtained from Airgas, Inc. (Buffalo, NY). Poly(PFMD) and poly(PFMMD) were synthesized by radical polymerization of perfluoro-2-methylene-1,3-dioxolane (PFMD) and perfluoro-2-methylene-4-methyl-1,3-dioxolane (PFMMD), respectively. The detailed synthesis procedures have been reported in the literature,19,20 and they are briefly described here. As shown in Figure 2, the perfluorodioxolane monomers were readily polymerized in HFB using perfluorobenzoyl peroxide (PFBP) as the initiator. The decomposition of PFBP takes place at 60−80 °C by a homolysis mechanism, resulting in the formation of pentafluorophenyl radicals, which initiate the polymerization and end up as structural units at the polymer chain ends. Nonfluorinated initiators such as azobis(isobutyronitrile) (AIBN) cannot initiate polymerization or only give a trace amount of polymer because nonfluorinated radicals are B

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 3. (a) DSC thermograms and (b) WAXD patterns of the pristine and annealed films of poly(PFMD) and poly(PFMMD). The annealing was conducted at 160 °C for 72 h. (c) WAXD patterns of commercial perfluoropolymers. imately 2 g of the polymer film was placed in the sample cell and evacuated overnight. Then a known amount of gas was introduced into the sample cell from the charge cell. As the gas dissolved into the polymer samples, the pressure in the sample cell decreased before reaching equilibrium. Gas solubility (SA, cm3 (STP)/(cm3 atm)) was calculated using the following equation:

SA = CA /pA

both polymers have degradation temperatures above 350 °C, and thus, degradation does not occur within the temperature range of the DSC scans. Both pristine polymers are amorphous as indicated by the absence of a crystallization or melting peak. Poly(PFMMD) exhibits a Tg of 135 °C, which is higher than that of poly(PFMD) (111 °C) because the bulky −CF3 substituent in poly(PFMMD) increases chain rigidity. These results are consistent with literature data.20,27 However, such a difference can also be partially caused by the molecular weight difference in these two polymers. Tg increases with increasing molecular weight of poly(PFMMD),17,20 and the molecular weight of poly(PFMD) is unknown. The effect of thermal annealing at 160 °C (above Tg) under vacuum for 72 h on the polymer morphology was also investigated. Annealing induced crystallization in poly(PFMD), as indicated by the Tm at 199 °C. This behavior is consistent with the literature, where processing conditions and heat treatment can lead to amorphous or semicrystalline poly(PFMD).27 On the other hand, annealing does not affect poly(PFMMD), probably because the −CF3 substituent introduces irregularity in the polymer structure and thus disrupts crystallization. For the gas transport property studies, pristine polymer films were used to avoid the effect of crystallinity. Figure 3b presents WAXD patterns of the pristine and annealed polymer films. The sharp peaks in annealed poly(PFMD) confirm crystallization derived from the annealing process, which is consistent with the DSC result. The broad

(1) 3

where CA is the concentration of the sorbed gas (cm (STP)/(cm3 polymer)) in the polymer at the equilibrium pressure of pA (atm). The obtained gas solubility has an uncertainty of less than 10%, which is estimated using propagation of errors analysis.26 The solubility of He, H2, and N2 in the polymers is too low to determine using this method.24

■

RESULTS AND DISCUSSION Physical Properties of Poly(PFMD) and Poly(PFMMD) Films. The average molecular weight (Mv) of poly(PFMMD) was 1.0 × 106 g/mol, which was calculated using the Mark− Houwink equation: [η] = KMva. The values of K and a are calculated using the known [η] and Mv values of poly(PFMMD) from previous studies, where Mv was determined by the light scattering method.20 Because poly(PFMD) is not soluble in solvents, the average molecular weight cannot be determined using this method or nuclear magnetic resonance (NMR). Figure 3a presents the DSC thermograms of pristine and annealed films of poly(PFMD) and poly(PFMMD) from 50 to 300 °C. As shown in Figure S1 of the Supporting Information, C

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Table 1. Comparison of Physical Properties and Gas Separation Properties at 35 °C in Poly(PFMD), Poly(PFMMD), and Commercial Glassy Perfluoropolymers29 permeability (barrer)

selectivity

polymer

Tg (°C)

d-spacinga (Å)

FFV

He

H2

N2

CO2

He/CH4

H2/CO2

poly(PFMD) poly(PFMMD) poly(PFMMD)b Teflon AF1600 Hyflon AD80 Cytop

111 135

8.1 (5.7 and 2.3) 11 (5.2)

0.21 0.23

210 560 648

9.1 (2.3) 8.2 (4.0 and 2.3) 5.9 (2.3)

0.31 0.23 0.21

0.71 7.7 9.7 110 24 5.0

5.9 58 66 520 150 35

1650 280 327

162 134 108

50 240 237 550 210 59

8.4 4.1 3.6 1.1 1.4 1.7

430 170

36 84

The values correspond to the main peak and the minor peaks (in the parentheses). bFrom ref 30. Films were prepared by pressing at 250 °C and ∼600 psi. The permeation testing temperature was not reported. a

Figure 4. Effect of feed pressure on the pure-gas permeability of the freestanding films of (a) poly(PFMD) and (b) poly(PFMMD) at 35 °C.

0.23, respectively. Poly(PFMMD) has a higher FFV value than poly(PFMD) due to the bulky −CF3 substituent. Table 1 compares the Tg and FFV values of poly(PFMD) and poly(PFMMD) with those of commercial perfluoropolymers. Poly(PFMD) exhibits Tg and FFV values similar to Cytop, while poly(PFMMD) shows values similar to Hyflon AD80. Gas Permeability of Poly(PFMD) and Poly(PFMMD). Figures 4a and 4b present pure-gas permeability of CH4, N2, Ar, CO2, H2, and He as a function of feed pressure in poly(PFMD) and poly(PFMMD) at 35 °C, respectively. Gas permeability is independent of feed pressure, suggesting that the films are defect-free. The detailed results are recorded in Table 1, which also compares the pure-gas permeability of poly(PFMMD) in this study with that reported in the literature.30 The gas permeabilities and selectivities reported here are consistent with the literature values, although the temperature for the permeation measurement was not reported in the literature case. As shown in Table 1, poly(PFMMD) exhibits much higher gas permeability and lower He/CH4 and H2/CO2 selectivity than poly(PFMD), presumably because of the higher FFV in poly(PFMMD) and the associated weaker size-sieving ability. The freestanding films of poly(PFMMD) demonstrate a He/ CH4 selectivity of 280 and a H2/CH4 selectivity of 120, which are much higher than those reported in the TFC membranes at 22 °C (108 and 57, respectively).5 The discrepancy can be caused by the small pinholes in the thin selective layer (∼200 nm) in the TFC membranes or substrate resistance for permeation of fast gases like He and H2 in the TFC membranes. On the other hand, the poly(PFMMD) films show a CO2/CH4 selectivity of 29 and a N2/CH4 selectivity of

peaks in the pristine samples confirm their amorphous nature. Compared with poly(PFMD), the presence of the bulky −CF3 substituent in poly(PFMMD) shifts the major peak to the left, indicating a d-spacing increase from 8.1 Å in poly(PFMD) to 11 Å in poly(PFMMD) (calculated using Bragg’s law). As shown in Table 1, poly(PFMD) has another d-spacing values of 5.7 and 2.3 Å, while poly(PFMMD) exhibits another d-spacing value of 5.7 Å. Figure 3c shows the WAXD patterns for the commercial perfluoropolymers, and their d-spacing values are recorded in Table 1. In general, the d-spacing corresponding to the major peaks increases in the following order: Cytop < Hyflon AD80 < Teflon AF1600, which is also consistent with the increasing order of gas diffusivity in these polymers. The position of the main peak for Teflon AF1600 is also consistent with the literature.28 Interestingly, they all show a minor peak corresponding to a d-spacing of 2.3 Å, which is consistent with that in Nafion,10 although such small free volume elements may not be accessible for gas molecules because the smallest gas molecule is He with a kinetic diameter of 2.6 Å. Hyflon AD80 shows an additional peak with a d-spacing of 4.0 Å. FFV is an important parameter to correlate with gas diffusivity in polymers and can be calculated using eq 2: FFV = 1 − ρV0

(2)

where ρ is the density of the amorphous polymer and V0 is the specific occupied volume at 0 K. The V0 is calculated as 1.3 times the van der Waals volume, which is estimated by Bondi’s group contribution method. 31 Poly(PFMD) and poly(PFMMD) have a density of 2.013 and 1.959 g/cm3, respectively, which corresponds to FFV values of 0.21 and D

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 5. Comparison of pure-gas selectivity (a) between poly(PFMD) and Cytop and (b) between poly(PFMMD) and Hyflon AD80 as a function of the difference in kinetic diameter of gas pairs.

Figure 6. Pure-gas sorption isotherms for various gases in (a) poly(PFMD) and (b) poly(PFMMD) films at 35 °C.

Table 2. Comparison of Solubility Parameter (δP), Gas Solubility, and Diffusivity in Poly(PFMD), Poly(PFMMD), Cytop, and Hyflon AD80 at 10 atm and 35 °C

a

polymer

δPa (MPa0.5)

SCO2 (cm3 (STP)/(cm3 atm))

SCO2/SCH4

SCO2/SC2H6

DCO2 (10−7 cm2/s)

DCO2/DCH4

poly(PFMD) Cytop29,33 poly(PFMMD) Hyflon AD804

10.3

1.3 1.3 1.4 1.5

3.3 4.3 2.9 2.0

1.2 1.5 1.2 1.1

0.36 2.0 3.2 8.2

14 4.2 9.9 7.7

10.3 10.8

Solubility parameters of the polymers are estimated using a group contribution method.31

poly(PFMMD) and Hyflon AD80. As shown in Figure 5b, poly(PFMMD) exhibits higher gas selectivity than Hyflon AD80, despite the similar Tg and FFV, and the difference in the gas selectivity increases with increasing size difference of the gas pairs. Gas Solubility and Diffusivity in Poly(PFMD) and Poly(PFMMD). To understand the significant differences in gas selectivity for dioxolane-based perfluoropolymers and commercial ones, gas solubility was determined. Sorption isotherms for various pure gases at 35 °C are depicted in Figure 6. The isotherms are concave, which is characteristic of gas sorption in glassy polymers.32 Gas sorption data can be described using the dual-mode sorption model, and the details of the modeling are presented in the Supporting Information as well as the

3.9, which are comparable with those reported for the TFC membranes at 22 °C (26 and 3.4, respectively).5 Figure 5a compares the gas selectivity between poly(PFMD) and Cytop as a function of the difference in the kinetic diameter of the gas pair. Both polymers have similar Tg and FFV. The selectivity increases with increasing values of the kinetic diameter difference. The CO2/CH4 selectivity in both polymers appears to be higher than expected from the trends because fluorinated polymers are hydrocarbon-phobic and thus show high CO2/CH4 solubility selectivity.4,6,10 Poly(PFMD) exhibits much higher gas selectivity or size-sieving ability than Cytop. As the kinetic diameter difference of the gas pair increases, the advantage of the selectivity of poly(PFMD) over that of Cytop becomes more significant. This behavior is similar for E

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 7. Comparison of the gas separation properties of poly(PFMD) (◆) and poly(PFMMD) (▲) with the commercial perfluoropolymers (cPFPs, △) and other polymers in Robeson’s upper-bound plots. (a) He/CH4, (b) He/H2, (c) H2/CH4, (d) H2/CO2, (e) N2/CH4, and (f) CO2/CH4 at 25−35 °C.8,39

polymers toward C2H64,6 and a possible steric suppression of C2H6 sorption because C2H6 has a critical volume (Vc) of 147 cm3/mol compared with CO2 with a Vc of 94 cm3/mol. In general, the perfluorodioxolanes show gas solubility behavior consistent with their perfluorinated nature, suggesting that the differences in permeation properties relative to commercial perfluoropolymers must be mostly related to diffusion properties. Based on the solution-diffusion mechanism, gas diffusivity (DA, cm2/s) in polymers can be calculated using eq 3:

adjustable parameters (cf. Table S1) and gas solubility at infinite dilution (cf. Table S2). Gas solubility at 10 atm in both polymers is also calculated and recorded in Table 2. Interestingly, all of these perfluoropolymers (including poly(PFMD), Cytop, poly(PFMMD), and Hyflon AD80) show similar CO2 solubility, which is consistent with their similarity in solubility parameter, functional groups, and FFV. These perfluoropolymers also show CO2/C2H6 solubility greater than 1, despite the similar critical temperatures for these two gases (i.e., 304 K for CO2 and 305 K for C2H6). This result is ascribed to the unfavorable interactions of perfluorinated

DA = PA /SA F

(3) DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

performance for He/CH4, H2/CH4, and H2/CO2 compared to the commercially available perfluoropolymers. The dioxolane-based perfluoropolymers also exhibit superior N2/CH4 separation performance (as shown in Figure 7f), while their CO2/CH4 separation performance is below the upper bound. On the other hand, they show better CO2/CH4 separation properties than the commercial workhorse membrane material (i.e., cellulose acetates). For example, poly(PFMD) exhibits higher selectivity than cellulose acetates with comparable CO2 permeability, while poly(PFMMD) shows higher permeability and comparable selectivity. These separation properties could make them promising for CO2/CH4 separation in natural gas treatment, considering the intrinsic resistance of perfluoropolymers to hydrocarbon plasticization.4,29,38

Gas diffusivity is usually influenced by the polymer FFV and chain rigidity, and the diffusivity selectivity of a smaller molecule over a larger one often increases with decreasing FFV and increasing chain rigidity. Table 2 shows very interesting behavior for CO2 diffusivity and CO2/CH4 diffusivity selectivity in these dioxolane-based perfluoropolymers. First, the dioxolane-based perfluoropolymers show much lower CO2 diffusivity than the commercial counterparts with similar Tg and FFV. For example, CO2 diffusivity in Cytop is about 5.5 times higher than that of poly(PFMD). Second, the CO2 diffusivity of poly(PFMD) is 1 order of magnitude less than that of poly(PFMMD), while the CO2/CH4 diffusivity selectivity in poly(PFMD) is 80% higher than that in poly(PFMMD). These rather large diffusion differences occur even though the difference in FFV values for all of these polymers is less than 10%. This finding might be qualitatively explained as the difference in the size and distribution of the free volume elements in these polymers. There are numerous examples in the literature where the type and shape of free volume elements can dramatically impact gas transport properties. For example, hourglass-shaped free volume elements with ultramicroporous necks have been shown to result in both high gas diffusivity and diffusivity selectivity.34 We suspect that the dioxolane-based perfluoropolymers have larger microcavities and smaller microporous necks than their commercial analogues with similar FFV values. As shown in Table 1, poly(PFMD) shows a trimodal distribution of free volume with the d-spacing values of 8.1 (the main peak), 5.7, and 2.3 Å, while Cytop with the same FFV value shows dspacing values of 5.9 and 2.3 Å. Poly(PFMMD) shows dspacing values of 11 and 5.2 Å, while Hyflon AD80 with the same FFV shows the d-spacing values of 8.2, 4.0 and 2.3 Å. The dioxolane-based perfluoropolymers provide larger d-spacings (from the main peak) and thus lower density of such free volume elements, leading to lower CO2 diffusivity than the commercial perfluoropolymers. The higher CO2/CH4 diffusivity selectivity (or stronger size-sieving ability) in the dioxolanebased perfluoropolymers can be ascribed to the more dominant effect on gas diffusion of the ultramicroporous necks connecting the large free volume elements. This is also consistent with the fact that the He/gas and H2/CO2 selectivity is much higher in dioxolane-based perfluoropolymers than in Cytop or Hyflon AD80, as shown in Table 1. These results provide further evidence that gas diffusivity selectivity is governed not only by the overall FFV values but also by the free volume size, distribution, and pore connectivity in polymers,34−37 which, unfortunately, cannot be directly measured using WAXD or other techniques such as positron annihilation lifetime spectroscopy (PALS). Superior Gas Separation Properties of Poly(PFMD) and Poly(PFMMD). Figure 7 demonstrates the superior gas separation properties of poly(PFMD) and poly(PFMMD) for He/CH4, H2/CO2, H2/CH4, He/H2, CO2/CH4, and N2/CH4 in Robeson’s upper-bound plots.8 For He/CH4 and He/H2 separation, poly(PFMD) and poly(PFMMD) show performance above or near the upper bound due to their strong sizesieving ability and unfavorable interactions toward CH4 and H2.2,6 For H2/CH4 and H2/CO2 separation, poly(PFMD) and poly(PFMMD) demonstrate performance above the upper bound due to their strong size-sieving ability. These dioxolanebased perfluoropolymers also show superior separation

■

CONCLUSION In this study, we demonstrate that dioxolane-based perfluoropolymers (poly(PFMD) and poly(PFMMD)) show gas separation performance above or near the upper bounds in Robeson’s plots for He/gas and gas/CH4 separations. Poly(PFMMD) exhibits higher gas permeability and weaker sizesieving ability than poly(PFMD) due to the bulky −CF3 substituent on the dioxolane ring. The transport properties of these perfluorodioxolanes were also compared with commercial amorphous glassy perfluoropolymers, such as Cytop and Hyflon AD80, which have been used for practical membrane gas separations. Gas solubility data for the perfluorodioxolanes are consistent with their perfluorinated nature and similar to the commercial perfluoropolymers. This result indicates that the superior combinations of permeability and selectivity for the perfluorodioxolanes must be related to diffusion properties. The dioxolane-based perfluoropolymers have lower gas diffusivity and higher diffusivity selectivity (stronger size-sieving ability) than the commercial polymers, despite their similar FFV and Tg. This behavior can be ascribed to the larger microcavities and narrower pore connections in the dioxolanebased perfluoropolymers, as evidenced by the d-spacing values derived from the WAXD patterns. The superior separation properties of poly(PFMD) and poly(PFMMD) make them promising for practical applications such as He/CH4 and CO2/ CH4 separation.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.8b00273. Thermogravimetric analysis for poly(PFMD) and poly(PFMMD); gas permeability; modeling of gas sorption isotherms using the dual mode sorption model (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*(Y.O.) Tel +1-646-997-3638; e-mail [email protected]. *(H.L.) Tel +1-716-645-1856, fax +1-716-645-3822; e-mail haiqingl@buffalo.edu. ORCID

Haiqing Lin: 0000-0001-8042-154X Notes

The authors declare no competing financial interest. G

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

■

(19) Okamoto, Y.; Mikeš, F.; Yang, Y.; Koike, Y. The effect of fluorine substituents on the polymerization mechanism of 2methylene-1,3-dioxolane and properties of the polymer products. J. Fluorine Chem. 2007, 128 (3), 202−206. (20) Mikeš, F.; Yang, Y.; Teraoka, I.; Ishigure, T.; Koike, Y.; Okamoto, Y. Synthesis and characterization of an amorphous perfluoropolymer: poly(perfluoro-2-methylene-4-methyl-1,3-dioxolane). Macromolecules 2005, 38 (10), 4237−4245. (21) Huggins, M. L. The viscosity of dilute solutions of long-chain molecules. IV. Dependence on concentration. J. Am. Chem. Soc. 1942, 64 (11), 2716−2718. (22) Jansen, J. C.; Macchione, M.; Drioli, E. On the unusual solvent retention and the effect on the gas transport in perfluorinated Hyflon AD (R) membranes. J. Membr. Sci. 2007, 287 (1), 132−137. (23) Zhao, S.; Huang, K.; Lin, H. Impregnated membranes for water purification using forward osmosis. Ind. Eng. Chem. Res. 2015, 54 (49), 12354−12366. (24) Lin, H.; Freeman, B. D. Permeation and diffusion. In SpringerHandbook of Materials Measurement Methods; Czichos, H., Smith, L. E., Saito, T., Eds.; Springer: Leipzig, 2006; pp 371−387. (25) Lin, H.; Freeman, B. D. Gas solubility, diffusivity and permeability in poly(ethylene oxide). J. Membr. Sci. 2004, 239 (1), 105−117. (26) Bevington, P. R.; Robinson, D. K. Data Reduction and Error Analysis for the Physical Sciences, 2nd ed.; McGraw-Hill, Inc.: New York, 1992. (27) Mikeš, F.; Baldrian, J.; Teng, H.; Koike, Y.; Okamoto, Y. Characterization and properties of semicrystalline and amorphous perfluoropolymer: poly(perfluoro-2-methylene-1,3-dioxolane). Polym. Adv. Technol. 2011, 22 (8), 1272−1277. (28) Tavana, H.; Jehnichen, D.; Grundke, K.; Hair, M. L.; Neumann, A. W. Contact angle hysteresis on fluoropolymer surfaces. Adv. Colloid Interface Sci. 2007, 134-135, 236−248. (29) Merkel, T. C.; Pinnau, I.; Prabhakar, R. S.; Freeman, B. D. Gas and vapor transport properties of perfluoropolymers. In Materials Science of Membranes for Gas and Vapor Separation; Freeman, B. D., Yampolskii, Y. P., Pinnau, I., Eds.; Wiley: 2006; pp 251−270. (30) Niedzielski, E. L.; Crest, G.; Putnam, R. E. Diffusion process and apparatus. US 3,307,330, 1967. (31) Van Krevelen, D. W.; Te Nijenhuis, K. In Properties of Polymers, 4th ed.; Elsevier: Amsterdam, 2009; Chapter 7, pp 189−227. (32) Kanehashi, S.; Nagai, K. Analysis of dual-mode model parameters for gas sorption in glassy polymers. J. Membr. Sci. 2005, 253 (1), 117−138. (33) Merkel, T. C.; Toy, L. G. Comparison of hydrogen sulfide transport properties in fluorinated and nonfluorinated polymers. Macromolecules 2006, 39 (22), 7591−7600. (34) Park, H. B.; Jung, C. H.; Lee, Y. M.; Hill, A. J.; Pas, J. P.; Mudie, S. T.; Van Wagner, E.; Freeman, B. D.; Cookson, D. J. Polymers with cavities tuned for fast selective transport of small molecules and ions. Science 2007, 318, 254−258. (35) Hofmann, D.; Entrialgo-Castano, M.; Lerbret, A.; Heuchel, M.; Yampolskii, Y. Molecular modeling investigation of free volume distributions in stiff chain polymers with conventional and ultrahigh free volume: Comparison between molecular modeling and positron lifetime studies. Macromolecules 2003, 36 (22), 8528−8538. (36) Belov, N. A.; Zharov, A. A.; Shashkin, A. V.; Shaikh, M. Q.; Raetzke, K.; Yampolskii, Y. P. Gas transport and free volume in hexafluoropropylene polymers. J. Membr. Sci. 2011, 383 (1−2), 70− 77. (37) Jansen, J. C.; Macchione, M.; Tocci, E.; De Lorenzo, L.; Yampolskii, Y. P.; Sanfirova, O.; Shantarovich, V. P.; Heuchel, M.; Hofmann, D.; Drioli, E. Comparative study of different probing techniques for the analysis of the free volume distribution in amorphous glassy perfluoropolymers. Macromolecules 2009, 42 (19), 7589−7604. (38) Lin, H.; Yavari, M. Upper bound of polymeric membranes for mixed-gas CO2/CH4 separations. J. Membr. Sci. 2015, 475, 101−109.

ACKNOWLEDGMENTS We gratefully acknowledge the financial support of this work by the U.S. National Science Foundation (NSF) under the CAREER Award No. 1554236 and the NSF STTR Award No. IIP-1632229.

■

REFERENCES

(1) Galizia, M.; Chi, W. S.; Smith, Z. P.; Merkel, T. C.; Baker, R. W.; Freeman, B. D. 50th anniversary perspective: polymers and mixed matrix membranes for gas and vapor separation: A review and prospective opportunities. Macromolecules 2017, 50 (20), 7809−7843. (2) Smith, Z. P.; Tiwari, R. R.; Dose, M. E.; Gleason, K. L.; Murphy, T. M.; Sanders, D. F.; Gunawan, G.; Robeson, L. M.; Paul, D. R.; Freeman, B. D. Influence of diffusivity and sorption on helium and hydrogen separations in hydrocarbon, silicon, and fluorocarbon-based polymers. Macromolecules 2014, 47 (9), 3170−3184. (3) Merkel, T. C.; Bondar, V.; Nagai, K.; Freeman, B. D.; Yampolskii, Y. P. Gas sorption, diffusion, and permeation in poly(2,2-bis(trifluoromethyl)-4,5-difluoro-1,3-dioxole-co-tetrafluoroethylene). Macromolecules 1999, 32 (25), 8427−8440. (4) Prabhakar, R. S.; Freeman, B. D.; Roman, I. Gas and vapor sorption and permeation in poly(2,2,4-trifluoro-5-trifluoromethoxy1,3-dioxole-co-tetrafluoroethylene). Macromolecules 2004, 37 (20), 7688−7697. (5) Fang, M.; He, Z.; Merkel, T.; Okamoto, Y. High-performance perfluorodioxolane copolymer membranes for gas separation with tailored selectivity enhancement. J. Mater. Chem. A 2018, 6, 652−658. (6) Yavari, M.; Okamoto, Y.; Lin, H. The role of halogens in polychlorotrifluoroethylene (PCTFE) in membrane gas separations. J. Membr. Sci. 2018, 548, 380−389. (7) Alentiev, A. Y.; Yampolskii, Y. P.; Shantarovich, V. P.; Nemser, S. M.; Plate, N. A. High transport parameters and free volume of perfluorodioxole copolymers. J. Membr. Sci. 1997, 126 (1), 123−132. (8) Robeson, L. M. The upper bound revisited. J. Membr. Sci. 2008, 320 (1−2), 390−400. (9) Park, H. B.; Kamcev, J.; Robeson, L. M.; Elimelech, M.; Freeman, B. D. Maximizing the right stuff: The trade-off between membrane permeability and selectivity. Science 2017, 356 (6343), eaab0530. (10) Mukaddam, M.; Litwiller, E.; Pinnau, I. Gas sorption, diffusion, and permeation in Nafion. Macromolecules 2016, 49 (1), 280−286. (11) Omidvar, M.; Nguyen, H.; Liu, J.; Lin, H. Sorption enhanced membrane materials for gas separation: A road less traveled. Curr. Opin. Chem. Eng. 2018, 20, 50−59. (12) Tiwari, R. R.; Smith, Z. P.; Lin, H.; Freeman, B. D.; Paul, D. R. Gas permeation in thin films of “high free-volume” glassy perfluoropolymers: Part I. Physical aging. Polymer 2014, 55 (22), 5788−5800. (13) Yavari, M.; Le, T.; Lin, H. Physical aging of glassy perfluoropolymers in thin film composite membranes. Part I. Gas transport properties. J. Membr. Sci. 2017, 525, 387−398. (14) Yavari, M.; Maruf, S. H.; Ding, Y. F.; Lin, H. Physical aging of glassy perfluoropolymers in thin film composite membranes. Part II. Glass transition temperature and the free volume model. J. Membr. Sci. 2017, 525, 399−408. (15) Lin, H.; Zhou, M.; Ly, J.; Vu, J.; Wijmans, J. G.; Merkel, T. C.; Jin, J.; Haldeman, A.; Wagener, E. H.; Rue, D. Membrane-based oxygen-enriched combustion. Ind. Eng. Chem. Res. 2013, 52 (31), 10820−10834. (16) Huang, Y.; Baker, R. W.; Wijmans, J. G. Perfluoro-coated hydrophilic membranes with improved selectivity. Ind. Eng. Chem. Res. 2013, 52 (3), 1141−1149. (17) Okamoto, Y.; Du, Q. M.; Koike, K.; Mikes, F.; Merkel, T. C.; He, Z. J.; Zhang, H.; Koike, Y. New amorphous perfluoro polymers: Perfluorodioxolane polymers for use as plastic optical fibers and gas separation membranes. Polym. Adv. Technol. 2016, 27 (1), 33−41. (18) Okamoto, Y.; Zhang, H.; Mikes, F.; Koike, Y.; He, Z. J.; Merkel, T. C. New perfluoro-dioxolane-based membranes for gas separations. J. Membr. Sci. 2014, 471, 412−419. H

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules (39) Freeman, B. D. Basis of permeability/selectivity tradeoff relations in polymeric gas separation membranes. Macromolecules 1999, 32 (2), 375−380.

I

DOI: 10.1021/acs.macromol.8b00273 Macromolecules XXXX, XXX, XXX−XXX