- Email: [email protected]
BPDA-DAM hollow fiber membranes at sub-ambient temperatures
Author’s Accepted Manuscript Post-combustion carbon dioxide capture via 6FDA/BPDA-DAM hollow fiber membranes at sub-ambient temperatures Lu Liu, Wulin Qiu, Edgar S. Sanders, Canghai Ma, William J. Koros www.elsevier.com/locate/memsci
PII: DOI: Reference:
S0376-7388(16)30155-7 http://dx.doi.org/10.1016/j.memsci.2016.03.027 MEMSCI14365
To appear in: Journal of Membrane Science Received date: 11 November 2015 Revised date: 4 March 2016 Accepted date: 13 March 2016 Cite this article as: Lu Liu, Wulin Qiu, Edgar S. Sanders, Canghai Ma and William J. Koros, Post-combustion carbon dioxide capture via 6FDA/BPDADAM hollow fiber membranes at sub-ambient temperatures, Journal of Membrane Science, http://dx.doi.org/10.1016/j.memsci.2016.03.027 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.
Post-combustion carbon dioxide capture via 6FDA/BPDA-DAM hollow fiber membranes at sub-ambient temperatures Lu Liua, Wulin Qiu a, Edgar S. Sandersb, Canghai Mab, William J. Korosa* a School of Chemical & Biomolecular Engineering, Georgia Institute of Technology, 311 Ferst Drive, Atlanta, GA 30332-0100, United States b
AirLiquide, Delaware Research and Technology Center, 200 GBC Drive, Newark, DE 19702
*
Email: [email protected] Phone: 404-385-2845 Fax: 404-385-2683
Abstract Post-combustion carbon dioxide capture has the greatest near-term potential compared to other technological pathways to mitigate CO2 emissions. The hybrid membrane and cryogenic distillation process developed at American AirLiquide allows sub-ambient temperature (-50 °C to -20 °C) CO2 capture to be cost effective. In previous work, Matrimid® hollow fiber membranes with so-called “nodular” selective layers were successfully formed and showed a combination of high selectivity and high permeance at temperatures below -20 °C. Despite good performance, considering the low CO2 concentration and large volume of flue gas to be treated, the permeance of the “nodular” Matrimid® fibers still required improvement. In this work, a new membrane material with higher fractional free volume than Matrimid®, 6FDA/BPDA-DAM (1:1), was investigated for sub-ambient temperature (SAT) CO2 capture. Defect free 6FDA/BPDA-DAM hollow fiber membranes were successfully formed; however, the bare fibers lost selectivity at temperatures below 0 °C, possibly due to the development of minor defects in the thin selective layer. This problem was corrected by post-treatment with polydimethyl siloxane (PDMS). The PDMS post-treated 6FDA/BPDA-DAM hollow fiber membranes displayed much higher permeance than the Matrimid® fibers with a high selectivity. The high selectivity is presumably due to additional polymer backbone rigidity at SAT conditions combined with possible local chain orientation; and the high permeance is believed to be a result of hindered packing of the polymer backbone. Keywords: Carbon dioxide capture; CO2/N2 separation; Hollow fiber membrane; Sub-ambient temperature; 6FDA/BPDA-DAM 1. Introduction There is growing concern that human-related carbon dioxide emissions are contributing to a warming world [1-3]. Over the past century, climate scientists observed a dramatic increase in the atmospheric carbon dioxide concentration, from a pre-industrial value of about 280 ppm to 394 ppm in 2012 [4]. If no carbon dioxide emission mitigation measure is carried out, this value is predicted to increase to 540-970 ppm by 2100 [5]. Among the many human activities that produce carbon dioxide, the combustion of fossil fuels (coal, oil and natural gas) is the largest source of CO2 emissions. In 2011, about 82 % of the global total primary energy supply was 1
generated by the combustion of fossil fuels [4]. Compared to oil and gas, coal provides the lowest cost method of producing electricity and is relatively abundant. In the last ten years, the emissions of CO2 from coal combustion increased significantly, which surpassed oil to become the largest source of CO2 emissions [4]. A recent study at MIT concluded that coal consumption will continue to increase into the foreseeable future [6]; therefore, it is important to control CO2 emissions from coal combustion. In previous work, we investigated post-combustion carbon dioxide capture via Matrimid® hollow fiber membranes at SAT conditions [7]. Asymmetric Matrimid® fibers with so-called “nodular” selectivity layers were post-treated with polydimethylsiloxane (PDMS) to provide a combination of high permeance and high selectivity at temperatures between -50 °C and -20 °C. Although cooling the flue gas to temperatures below -20 °C requires considerable energy input, the hybrid membrane and cryogenic distillation process developed at American AirLiquide makes the capture of carbon dioxide at such low temperatures cost effective. A detailed description of the process can be found in previous work [8, 9]. The coupling of the unique high performing membranes with cryogenic processing technology ensures efficient capture of carbon dioxide. The low CO2 concentration and large volume of flue gas to be treated make membrane productivity very important, so higher CO2 permeance is an attractive target of this work. In this work, a new membrane material, 6FDA/BPDA-DAM (1:1), was investigated for SAT carbon dioxide capture. The full name of 6FDA/BPDA-DAM is 4,4’(hexafluoroisopropylidene) diphthalic anhydride (6FDA)/ 3,3’-4,4’-biphenyl tetracarboxylic acid dianhydride (BPDA) - 2,4,6-trimethyl-1,3-phenylene diamine (DAM). The high free volume, rigid 6FDA/BPDA-DAM polyimide has been well characterized by previous researchers in terms of its structure and ambient temperature intrinsic gas separation properties [10-12]. Moreover, 6FDA/BPDA-DAM-derived carbon molecular sieve (CMS) hollow fiber membranes show attractive gas separation properties [13-16] versus less rigid polymer precursors (e.g. Matrimid®). However, so far, there is no report on the gas separation properties of the 6FDA/BPDA-DAM fibers at sub-ambient temperatures. The high fractional free volume of 6FDA/BPDA-DAM contributes to higher membrane permeability than Matrimid® at room and high temperatures [12-14], and the same trend was expected at SAT conditions. In the current work, 6FDA/BPDA-DAM (1:1) was made into a hollow fiber configuration, and its gas separation properties at SAT conditions were investigated and compared with the high performing “nodular” Matrimid® hollow fiber membranes. 2. Experimental 2.1. Materials The asymmetric hollow fiber membranes studied in the current work was formed from the polyimide 6FDA/BPDA-DAM (1:1). It was synthesized via a two-step polycondensation reaction described in other work [12]. The first step involves the reaction of stoichiometric amounts of the dianhydrides 6FDA and BPDA in 1-methyl-2-pyrrolidinone (NMP) at temperatures around 5 °C to produce a high molecular weight polyamic acid; and the second step is an imidization reaction which closes the imide ring via dehydration. The 6FDA/BPDA-DAM (1:1) polymer was dried in a vacuum oven at 100 °C overnight before use. The structure of its 2
repeating unit is shown in Fig.1 and its physical properties are shown in Table 1. The pure gas transport properties of 6FDA/BPDA-DAM dense film membranes at 35 °C were studied by previous researchers and are shown in Table 2.
Fig.1. The repeating unit of 6FDA/BPDA-DAM. The ratio of X to Y is 1:1.
Table 1 Physical properties of 6FDA/BPDA-DAM (1:1) and Matrimid® [11]. Fractional Free Polymer Density (g/cm3) Tg (°C) Volume (FFV) Matrimid® 1.25 0.110 305 6FDA/BPDA-DAM (1:1)
1.32
0.145
424
Table 2 Pure gas transport properties of 6FDA/BPDA-DAM (1:1) dense films at feed pressures < 100 psia and 35 °C [11, 17] Permeability (Barrer) Material
6FDA/BPDA-DAM
1 Barrer 11010
Selectivity
O2
CO2 C2H4 C3H6
O2 N2
CO2 N2
CO2 CH 4
C2 H 4 C2 H 6
C3 H 6 C3 H 8
64
309
4.1
19.8
22
3.3
13
46
11.8
cm3 ( STP) cm cm2 s cmHg
2.2. Formation of asymmetric hollow fiber membranes The asymmetric hollow fiber membranes used in this work was formed via the wellknown “dry-jet/wet-quench” spinning technique. Detailed description of this process can be found in our earlier papers - based on simultaneous co-extrusion of the dope and bore fluid through a spinneret [7, 18-20]. Once extruded, the nascent fiber was drawn in an air gap, where solvent and/or non-solvent evaporation occurred at the outermost portion of the fiber to form a nascent ultra-thin selective layer. The nascent fiber then entered a water quench bath where phase separation occurred and a solid fiber was formed. The solid fibers were collected and immersed in deionized (DI) water for 4 days, with the water being changed daily. This promoted the removal of most of the trace residual solvent and non-solvent from the fibers. The fibers were 3
then sequentially soaked in low surface tension methanol and hexane for 1 hour, with the solvents being changed every 20 minutes. After solvent exchange, the fibers were air-dried in a fume hood for 1 hour and dried under vacuum at 75 °C for 3 hours. 2.3. Permeation Both pure and mixed gas permeation experiments in the current work were carried out in a constant pressure system, and a countercurrent flow configuration was used. For pure gas permeation testing, feed was introduced at the bore side of the hollow fiber membranes, and the permeate was collected at the shell side. The permeate flow rate was measured via a bubble flowmeter. A typical hollow fiber module in this work contains 2 fibers and has an effective length of around 20 cm. For mixed gas permeation, a bore side feed flowrate was maintained to prevent concentration polarization and maintain an essentially constant feed. In this work, the majority of the mixed gas permeation testing was carried out at temperatures from -50 °C to 35 °C. A BTZ475 benchtop temperature chamber (ESPEC North America, INC., Hudsonville, MI), which accurately controlled temperature within the range of -70 °C to 180 °C with ± 0.5 °C fluctuation, was used to contain the modules. The permeate flow rate was measured by a bubble flow meter and the permeate composition was analyzed by a gas chromatograph (Bruker 430-GC, Agilent Technologies, CA). The retentate flow rate was controlled by a needle valve and was measured by a digital mass flowmeter (FMA Series, Omega Engineering, Inc.) to provide low stage cuts, typically below 1%. 2.4. Pressure decay sorption The pressure decay method was used in the current work to characterize gas sorption by the polymer. The design considerations of a pressure-decay apparatus has been described earlier [21], and some modifications were made to construct the pressure-decay apparatus for sorption testing at cold conditions. The same BTZ-475 benchtop temperature chamber (ESPEC North America, INC., Hudsonville, MI) as used in the permeation system was selected to contain the pressure decay apparatus. The apparatus consists of two main compartments: the reservoir and the sample cell. The polymers were placed in the sample cell, and pressure transducers were used to monitor the pressures in each compartment. All the valves within in the temperature chamber were pneumatically-actuated, and were controlled by National Instruments LabVIEW 8.6 code. The equilibrium sorbed concentration of the gas was calculated through a standard mole balance. 2.5. Scanning Electron Microscopy (SEM) The morphology of the asymmetric hollow fiber membranes was examined by scanning electron microscopy (SEM). A high resolution SEM (LEO 1530) was used in this work. The LEO 1530 was equipped with a thermally assisted field emission gun, and an 8 kV operation voltage was found to display a clear image. In the preparation of samples, dried fibers were soaked in hexane for ~ 1 minute and then quickly transferred into liquid nitrogen. The fibers were shear fractured in liquid nitrogen and then mounted vertically on sample holders. Prior to examination, samples were sputter coated with gold for 30 seconds. 4
3. Results and discussion 3.1. Formation of 6FDA/BPDA-DAM hollow fiber membranes The formation of 6FDA-based hollow fiber membranes has been studied by several researchers in the past. Wallace [22] and Chen [23] successfully formed defect free asymmetric hollow fiber membranes from 6FDA-DAM/DABA (4:1) and 6FDA-DAM/DABA (3:2), respectively. Weinberg [24] studied the formation of 6FDA/BPDA-DAM hollow fiber membranes using dimethyl sulfoxide (DMSO) as one of the solvents, but no defect-free selective layer formation was reported. Recently, Xu [17, 25] successfully formed defect free 6FDA-DAM and 6FDA/BPDA-DAM asymmetric hollow fiber membranes. In this work, 6FDA/BPDA-DAM (1:1) with a somewhat lower molecular weight (123 kDa) than Xu’s polymer (160 kDa) was used as the membrane material; however, this had negligible impact on spinning. This polymer has a polydispersity index of 2.0 (Table 3). Table 3 Molecular weight and polydispersity index of 6FDA/BPDA-DAM (1:1) used in the current work Polymer
Mw (kDa)
PDI (Mw/Mn)
6FDA/BPDA-DAM (1:1)
123
2.0
Compared to the spinning of Matrimid® hollow fiber membranes, 6FDA/BPDA-DAM requires more non-solvent for phase separation at equivalent polymer concentration. Reflected on the ternary phase diagram, the binodal line for 6FDA/BPDA-DAM is closer to the nonsolvent vertex (Figure 2). Slower phase separation tends to occur for 6F-based polymers compared to Matrimid®, so lithium nitrate (LiNO3) was introduced to the dope as an additional non-solvent and pore forming component. Lithium nitrate complexes with the solvent NMP molecules in the dope and the complex dissociates in the water quench bath and thus accelerates phase separation [17]. A 6.5 wt. % lithium nitrate level was found in our earlier work to be an appropriate concentration [17, 26, 27], and the same concentration was used for the current work. N-methyl-2-pyrrolidone (NMP) and tetrahydrofuran (THF) were used as the solvents, and ethanol (EtOH) was selected as the non-solvent. A dope’s viscosity and ability to phase separate are two critical parameters for the formation of desirable hollow fiber membranes, which are determined by the initial dope composition. Adequate viscosity allows the dope to tolerate the rigorous extensional stresses during fiber spinning, and rapid phase separation promotes the formation of open pores in the substructure [28]. A dope’s initial composition should be close to the binodal line (the boundary between the one-phase and two-phase regions) on the ternary phase diagram. The binodal line in this work was determined via the “cloud point” technique [29, 30]. This technique involves the preparation of a series of dope samples with different compositions. In the current work, three polymer concentrations were selected: 16 wt.%, 20 wt.% and 25 wt.%. At each polymer concentration, dope samples with increasing non-solvent contents (and accordingly decreasing solvent contents) were prepared and well mixed. The dope samples with a composition sitting in 5
the one-phase region were transparent, whereas those with a composition sitting in the two-phase region appeared cloudy. The composition on the phase boundary was the “cloud point”, as is shown by the open circles in Fig. 2. All the cloud points were connected to form an approximate binodal line (the solid line in Fig. 2). The considerable difference in the binodal for this 6FDAbased polymer versus the simpler Matrimid® polymer (the dashed line in Fig. 2) is obvious. The solid red point on the ternary phase diagram is the initial dope composition used in the current work, and the detailed composition is listed in Table 4.
Fig.2. Initial dope composition and binodal line for the 6FDA/BPDA-DAM (1:1) used in the current work. The solid line is the binodal for 6FDA/BPDA-DAM, and the dashed line is the binodal for Matrimid®[17]. Open circles represent cloud points for 6FDA/BPDA-DAM in the current work, and the solid red point represents the initial dope composition. After two spinning attempts, defect free 6FDA/BPDA-DAM hollow fiber membranes were successfully formed. The spinning conditions for the successful attempt are listed in Table 5. A high spinneret temperature (70 °C) was used to promote the evaporation of volatile solvents and non-solvents in the air gap, and a high quench temperature (50 °C) was used to accelerate phase separation. The SEM images of the resultant fiber’s cross section and selective layer are shown in Fig. 3.
6
Table 4 Initial dope composition for the spinning of 6FDA/BPDA-DAM (123 kDa) hollow fiber membranes Component
wt.%
6FDA/BPDA-DAM
20
NMP
47.5
THF
10
Ethanol
16
LiNO3
6.5
Table 5 Spinning conditions for the formation of defect free 6FDA/BPDA-DAM hollow fiber membranes Spinning parameter
Value
Dope flow rate
180 ml/h
Bore flow rate
60 ml/h
Bore fluid composition
85%/15% NMP/H2O
Take-up rate
30 m/min
Quench temperature
49 °C
Spinneret temperature
70 °C
Air gap height Room temperature Relative humidity
10, 15, 20 cm 25 °C 33-36 %
7
Fig.3. SEM images for the cross-section and selective layer of the defect free 6FDA/BPDADAM hollow fiber membranes. Skin integrity of the defect free 6FDA/BPDA-DAM hollow fiber membranes was confirmed by permeating pure O2, N2 and CO2 at 35 °C. Feed pressure for O2 and N2 was 100 psig; and for CO2, 50 psig. The O2/N2 selectivity was 3.93 ± 0.11 (Table 6), which was more than 90 % of the 6FDA/BPDA-DAM dense film value, suggesting the hollow fiber membranes were essentially defect free. (The O2/N2 selectivity of 6FDA/BPDA-DAM dense film at 35 °C is 4.1 [10]). An asymmetric hollow fiber membrane is conventionally considered to be defect free if its selectivity for a given gas pair is higher than 90 % of the polymer’s intrinsic dense film value [20]. Table 6 Pure gas permeation results for defect free 6FDA/BPDA-DAM hollow fiber membranes at 35 °C. Feed pressure for O2 and N2 was 100 psig; for CO2, 50 psig. The feed was introduced at the bore side.
PO2 l
(GPU)
82.8 ± 4
PN 2 l
(GPU)
21.1 ± 0.5
PCO2 l
(GPU)
498 ± 8
O2 / N2
CO2 / N2
3.93 ± 0.11
23.7 ± 0.1
3.2. Mixed gas permeation To investigate the separation properties of defect free 6FDA/BPDA-DAM hollow fiber membranes at cold conditions, mixed gas permeation measurement was carried out. A mixture of 20 mol% CO2/80 mol% N2 was introduced at the bore side of the fibers at 100 psig. In the permeation calculation, fugacity was used instead of partial pressure to account for the nonidealities of gases, and pressure drop along the fibers were also considered. Starting from 35 °C, the temperature was reduced step by step, with the feed continuously purging the module. As 8
70
3500
60
3000
CO2 Permeance (GPU)
CO2/N2 Selectivity
temperature was decreased from 35 °C to 0 °C, an increase in selectivity and decrease in permeance was observed (Fig. 4), as was expected due to the activated nature of permeation. The higher CO2/N2 selectivity at lower temperature was attributed to the lower permeation activation energy for CO2 than for N2 [18, 31]. However, as temperature was decreased further, the membrane lost selectivity and a dramatic increase in permeate flow rate was observed. This experiment was repeated 4 times, and the same results were obtained. At first, it was suspected that water might be the cause for the sudden loss of selectivity. Crystallization of residue water in the fiber backbone might lead to change of overall volume, which might subsequently lead to rupture of fiber. However, the thermogravimetric analysis (TGA) results confirmed that there was minimal amount of residue water present in the fiber, as is shown in Fig. 5. No water peak was observed. The weight loss between 80 °C and 120 °C is less than 0.5%, indicating the amount of water in the 6FDA/BPDA-DAM fibers is negligible. As the permeation temperature was brought back from -40 °C to 35 °C, the membrane regained some selectivity. However, the new selectivity was slightly lower than the original value (~ 12 % lower), and a higher membrane permeance was observed (~ 28 % higher). This unusual phenomenon suggested the thin selective layer of the membranes might have developed some defects at temperatures below 0 °C, possibly due to stresses on the increasingly rigid polymer chains as temperature decreased [32, 33]. Similar phenomena were observed for defect free Matrimid® hollow fiber membranes, which started to lose selectivity at temperatures below 20 °C.
50 40 30 20 10 0 -50 -40 -30 -20 -10 0
10 20 30 40
2500 2000 1500 1000 500 0 -50 -40 -30 -20 -10 0
10 20 30 40 o
o
Temperature ( C)
Temperature ( C)
(a) (b) Fig.4. Mixed gas permeation results of defect free 6FDA/BPDA-DAM hollow fiber membranes at different temperatures. The feed was a mixture of 20 mol% CO2/80 mol% N2, and was introduced at the bore side at 100 psig.
9
80 70
0.6
6FDA/BPDA-DAM(1:1) fiber TGA heating at 10 K/min in N2
0.4
60 0.2 50 40
o
Residual weight (wt%)
0.8
90
Derivative weight (% / C)
1.0
100
0.0 100 200 300 400 500 600 700 800 900 o
Temperature C
Fig.5. TGA results of 6FDA/BPDA-DAM (1:1) fibers. Sample was heated directly from 30 °C to 900 °C in a N2 atmosphere. The heating rate was 10 K/min.
3.3. Performance of PDMS post-treated 6FDA/BPDA-DAM membranes PDMS post-treatment is known to be effective in plugging pinhole defects in the membrane’s selective layers and thus enhance membrane performance [7, 18]. We wondered whether such a treatment could help prevent the 6FDA/BPDA-DAM hollow fiber membranes from losing selectivity at cold conditions. In this work, a 2 wt.% Sylgard® 184 (a two-part, silicon elastomer kit [34]) was dissolved in heptane to prepare the PDMS solution. To achieve a desirable PDMS molecular weight, the solution was heated at 100 °C in an oil bath for 5 hours with stirring until a slight increase in viscosity of the solution was observed. This increase in viscosity reflected chain growth and incipient network formation of PDMS chains, which were effective in caulking minor defects in the membranes [35]. Fibers in a module were shell-side soaked in this mixture for 30 minutes, after which the solution was drained. Then the module was dried in a fume hood for 24 hours at room temperature, followed by curing at 78 °C for 2 hours under vacuum. The PDMS post-treated 6FDA/BPDA-DAM hollow fiber membranes were tested at subambient temperatures. The same testing conditions as described in Section 3.2 were used. The 20 mol% CO2/80 mol% N2 was introduced at the bore side of the fibers at 100 psig. Interestingly, the PDMS post-treated 6FDA/BPDA-DAM fibers didn’t lose selectivity when the temperature was dropped below 0 °C, as is shown in Fig. 6. The CO2/N2 selectivity kept increasing with the decrease of temperature over a range of -50 °C to 35 °C (Fig. 6a). At -50 °C, the CO2/N2 selectivity was 214 ± 19, more than 10 times higher than the selectivity at 35 °C (20.9 ± 0.6). The CO2 permeance decreased slightly at first with decreasing temperature, and then the trend was reversed (Fig. 6b). The subsequent increase in CO2 permeance was believed to largely
10
reflect the stronger increase in CO2 solubility versus N2 at lower temperatures. This factor mitigates CO2 permeance reduction, despite decreased diffusivity.
CO2/N2 Selectivity
200 150 100 50 0 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50
2500
CO2 Permeance (GPU)
PDMS post-treated Un-treated
250
PDMS post-treated Un-treated
2000 1500 1000 500
0 -60 -50 -40 -30 -20 -10 0 10 20 30 40 o
o
Temperature ( C)
Temperature ( C)
.
(a) (b) Fig.6. Comparison of the performance of PDMS post-treated and un-treated defect free 6FDA/BPDA-DAM hollow fiber membranes at different temperatures. The feed was a mixture of 20 mol% CO2/80 mol% N2, and was introduced at the bore side at 100 psig.
The PDMS post-treatment was found to decrease the rigidity of 6FDA/BPDA-DAM (1:1) fibers. Fig. 7 shows the results of dynamic mechanical analysis (DMA) of un-treated and PDMS post-treated 6FDA/BPDA-DAM fibers. The PDMS post-treated fibers showed lower storage modulus and lower stiffness compared to the un-treated sample, indicating better mechanical strength. The PDMS post-treatment was found earlier to prevent defect free Matrimid® hollow fiber membranes from losing selectivity at sub-ambient temperatures as well. To sum up, in order for 6FDA/BPDA-DAM and Matrimid® hollow fiber membranes to achieve satisfying performance at SAT conditions, PDMS post-treatment is necessary. 700
550 500 450 400 350
Un-treated fiber PDMS treated fiber
1250
600 Stiffness (N/m)
Storage Modulus (MPa)
1300
Un-treated fiber PDMS treated fiber
650
1200 1150 1100 1050
40
60
80
100
120
140
40
o
60
80
100
120
140
o
Temperature ( C)
Temperature ( C)
Fig.7. DMA results of un-treated and PDMS post-treated 6FDA/BPDA-DAM (1:1) fibers. 11
3.4. Comparison of 6FDA/BPDA-DAM and Matrimid® hollow fiber membranes The performance of the PDMS post-treated 6FDA/BPDA-DAM hollow fiber membranes were also compared with the high performing “nodular” Matrimid® fibers developed in our previous work [7] as well as the defect free Matrimid® fibers. As discussed above, PDMS posttreatment is necessary for 6FDA/BPDA-DAM and Matrimid® fibers to achieve satisfying performance. Therefore, all fibers compared here were post-treated with PDMS. It was interesting to discover that the 6FDA/BPDA-DAM fibers showed much higher CO2 permeance than the Matrimid® fibers at SAT conditions; and at the same time, a high CO2/N2 selectivity was maintained (Fig. 8).
CO2/N2 Selectivity
250
Defect free 6F ® Nodular Matrimid ® Defect free Matrimid
200 150 100 50 0 -60 -50 -40 -30 -20 -10 0 10 20 30 40
700
CO2 Permeance (GPU)
300
o
600 500
Defect free 6F ® Nodular Matrimid ® Defect free Matrimid
400 300 200 100 0 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 o
Temperature ( C)
Temperature ( C)
(a) (b) Fig.8. Comparison of the performance of PDMS post-treated defect free 6FDA/BPDA-DAM and Matrimid® hollow fiber membranes at different temperatures. All fibers compared here are posttreated with PDMS. The feed was a mixture of 20 mol% CO2/80 mol% N2, and was introduced at the bore side at 100 psig.
The high CO2/N2 selectivity for 6FDA/BPDA-DAM fibers at SAT conditions was presumably due to additional polymer backbone rigidity coupled with possible local chain orientation; and the high CO2 permeance was believed to be due to higher CO2 solubility and higher CO2 diffusivity in the polymer compared to Matrimid®. The presence of bulky -CF3 groups in the rigid 6FDA/BPDA-DAM backbones hinders the packing of the polymer chains, leading to a larger Langmuir sorption capacity ( CH' ) than Matrimid®. The sorption of CO2 in 6FDA/BPDA-DAM at room and sub-ambient temperatures was carried out in the current work. The gas uptake was calculated by mole balance, and the non-ideality of CO2 was taken into account. Appropriate compressibility factors were used for the different operating temperatures, and fugacity was used instead of partial pressure. The CO2 sorption isotherms for 6FDA/BPDADAM at 35 °C and -40 °C are compared with the isotherms for Matrimid® (Fig. 9). At each temperature, a new sample was used to avoid the influence of polymer matrix swelling and conditioning due to the sorption of highly condensable CO2 at high partial pressures on measured gas uptake. For each set of data, two measurements were made, and a standard deviation of less 12
than 10% was observed. The data shown in Fig. 9 is the average of the two measurements. The discrete points are experimental data, and the solid lines are the fit of concentration versus fugacity data to the dual-mode sorption model. As is shown in Fig. 9, the gas uptake in 6FDA/BPDA-DAM was much higher than in Matrimid®, confirming the higher expected CO2 sorption in 6FDA/BPDA-DAM. The dual mode parameters derived from the sorption isotherms are listed in Table 7. The larger Langmuir sorption capacity ( CH' ) for CO2 in 6FDA/BPDA-
70
6FDA/BPDA-DAM ® Matrimid
60
CO2 Concentration (CCSTP/CC)
CO2 Concentration (CCSTP/CC)
DAM, a reflection of the amount of unrelaxed free volume in the samples being tested, is consistent with its higher solubility in the polymer. Since the selective layer thickness of the PDMS post-treated membranes cannot be determined reliably, actual diffusion coefficients cannot be determined. Nevertheless, the high free volume of 6FDA/BPDA-DAM may contribute to a larger number of transient gaps for CO2 molecules to jump through, resulting in higher CO2 diffusivity.
o
50
35 C
40 30 20 10 0
0
20
40
60
80
100
120
140
450
6FDA/BPDA-DAM ® Matrimid
400 350
o
- 40 C
300 250 200 150 100 50 0
0
10
20
Fugacity (psia)
30
40
50
60
70
Fugacity (psia)
Fig.9. The CO2 sorption isotherms for 6FDA/BPDA-DAM (1:1) at 35 °C and -40 °C as compared to Matrimid®. The discrete points are experimental data and the solid lines are the fit of concentration versus fugacity data to the dual-mode sorption model. The 6FDA/BPDA-DAM (1:1) samples for the sorption measurement has a molecular weight of 123 kDa.
Table 7 Dual-mode sorption parameters for CO2 in 6FDA/BPDA-DAM (1:1) and Matrimid® CH' (
T (°C)
35 -40
cm3 STP cm3 poly
)
6FDA/BPDA -DAM
Matrimid®
35.9 ± 1.2 112 ± 3
22.5 ± 2.5 47.4 ± 1.0
1 b psia 6FDA/BPDA Matrimid® -DAM
0.06 ± 0.01 0.86 ± 0.14
13
0.07 ± 0.01 0.82 ± 0.08
kD (
cm3 STP
cm3 poly psia
)
6FDA/BPDADAM
Matrimid®
0.19 ± 0.01 3.63 ± 0.06
0.14 ± 0.02 1.47 ± 0.03
4. Conclusions The rigid, packing inhibited 6FDA/BPDA-DAM (1:1) polyimide was investigated in this work for SAT conditions carbon dioxide capture. Defect free 6FDA/BPDA-DAM (1:1) hollow fiber membranes were successfully formed via the optimization of initial dope composition and spinning conditions. Mixed gas permeation experiments on the bare 6FDA/BPDA-DAM fibers showed a significant loss of selectivity and a dramatic increase in membrane permeance at temperatures below 0 °C. This was attributed to the development of minor defects in the membrane’s thin selective layer. However, it was interesting to discover that treating the 6FDA/BPDA-DAM fibers in a PDMS solution could correct this problem. The PDMS posttreated 6FDA/BPDA-DAM fibers showed increasing CO2/N2 selectivity as temperature was decreased from 35 °C to -50 °C. Compared to the high performing “nodular” Matrimid® hollow fiber membranes developed in a previous work, the 6FDA/BPDA-DAM fibers showed much higher permeance at SAT conditions; meanwhile, a high CO2/N2 selectivity was maintained. Considering the low CO2 concentration and large volume of flue gas to be treated, a higher membrane permeance will contribute to more efficient CO2 capture. The high permeance of 6FDA/BPDA-DAM hollow fiber membranes was attributed to the hindered packing of the polymer backbones, leading to higher CO2 solubility and diffusivity than the more densely packed Matrimid®.
Acknowledgement The authors would like to thank American AirLiquide for the funding of this project. References [1] I. P. O. C. Change, Climate change 2007: The physical science basis, Agenda, vol. 6, p. 333, 2007. [2] H. Yang, Z. Xu, M. Fan, R. Gupta, R. B. Slimane, A. E. Bland, et al., Progress in carbon dioxide separation and capture: A review, Journal of Environmental Sciences, vol. 20, pp. 14-27, 2008. [3] D. Aaron and C. Tsouris, Separation of CO2 from flue gas: a review, Separation Science and Technology, vol. 40, pp. 321-348, 2005. [4] CO2 Emissions From Fuel Combustion Highlights 2013 Edition, International Energy Agency, 2013. [5] J. T. Houghton, Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden, X. Dai, et al., Climate change 2001: the scientific basis vol. 881: Cambridge University Press Cambridge, 2001. [6] The Future of Coal - Options for a Carbon-Constrained World, MIT Interdisciplinary Study, 2007. [7] L. Liu, E. S. Sanders, S. S. Kulkarni, D. J. Hasse, and W. J. Koros, Sub-ambient temperature flue gas carbon dioxide capture via Matrimid® hollow fiber membranes, Journal of Membrane Science, vol. 465, pp. 49-55, 2014.
14
[8]
[9] [10] [11] [12] [13] [14]
[15]
[16]
[17] [18]
[19] [20]
[21]
[22]
[23] [24] [25]
D. J. Hasse, S. S. Kulkarni, E. S. Sanders Jr, J.-P. Tranier, and P. Terrien, Method of obtaining carbon dioxide from a carbon dioxide-containing gas mixture, ed: Google Patents, 2013. D. Hasse, S. Kulkarni, E. Sanders, E. Corson, and J.-P. Tranier, CO2 capture by subambient membrane operation, Energy Procedia, vol. 37, pp. 993-1003, 2013. K. M. Steel, Ph.D. thesis, Carbon membranes for challenging gas separations, The University of Texas at Austin, 2000. M. Rungta, Ph.D. thesis, Carbon molecular sieve dense film membranes for ethylene/ethane separations, Georgia Institute of Technology, 2012. M. Kiyono, Ph.D. thesis, Carbon molecular sieve membranes for natural gas separations, Georgia Institute of Technology, 2010. K. M. Steel and W. J. Koros, An investigation of the effects of pyrolysis parameters on gas separation properties of carbon materials, Carbon, vol. 43, pp. 1843-1856, 2005. D. Q. Vu, W. J. Koros, and S. J. Miller, High pressure CO2/CH4 separation using carbon molecular sieve hollow fiber membranes, Industrial & engineering chemistry research, vol. 41, pp. 367-380, 2002. M. Kiyono, P. J. Williams, and W. J. Koros, Effect of polymer precursors on carbon molecular sieve structure and separation performance properties, Carbon, vol. 48, pp. 4432-4441, 2010. L. Xu, M. Rungta, M. K. Brayden, M. V. Martinez, B. A. Stears, G. A. Barbay, et al., Olefins-selective asymmetric carbon molecular sieve hollow fiber membranes for hybrid membrane-distillation processes for olefin/paraffin separations, Journal of Membrane Science, vol. 423, pp. 314-323, 2012. L. Xu, Ph.D. thesis, Carbon molecular sieve hollow fiber membranes for olefin/paraffin separations, Georgia Institute of Technology, 2012. L. Liu, E. S. Sanders, J. R. Johnson, O. Karvan, S. Kulkarni, D. J. Hasse, et al., Influence of membrane skin morphology on CO2/N2 separation at sub-ambient temperatures, Journal of Membrane Science, vol. 446, pp. 433-439, 2013. D. T. Clausi and W. J. Koros, Formation of defect-free polyimide hollow fiber membranes for gas separations, Journal of Membrane Science, vol. 167, pp. 79-89, 2000. S. B. Carruthers, G. L. Ramos, and W. J. Koros, Morphology of integral‐skin layers in hollow‐fiber gas‐separation membranes, Journal of applied polymer science, vol. 90, pp. 399-411, 2003. W. J. Koros and D. Paul, Design considerations for measurement of gas sorption in polymers by pressure decay, Journal of Polymer Science: Polymer Physics Edition, vol. 14, pp. 1903-1907, 1976. D. W. Wallace, Ph.D. thesis, Crosslinked hollow fiber membranes for natural gas purification and their manufacture from novel polymers, The University of Texas at Austin, 2004. C.-C. Chen, Ph.D. thesis, Thermally crosslinked polyimide hollow fiber membranes for natural gas purification, Georgia Institute of Technology, 2011. M. G. Weinberg, Polyimide gas separation membranes for carbon dioxide enrichment, ed: Google Patents, 1993. L. Xu, C. Zhang, M. Rungta, W. Qiu, J. Liu, and W. J. Koros, Formation of defect-free 6FDA-DAM asymmetric hollow fiber membranes for gas separations, Journal of Membrane Science, vol. 459, pp. 223-232, 2014. 15
[26] [27]
[28]
[29]
[30]
[31] [32] [33] [34] [35]
I. C. Omole, Ph.D. thesis, Crosslinked polyimide hollow fiber membranes for aggressive natural gas feed streams, Georgia Institute of Technology, 2008. C.-C. Chen, W. Qiu, S. J. Miller, and W. J. Koros, Plasticization-resistant hollow fiber membranes for CO2/CH4 separation based on a thermally crosslinkable polyimide, Journal of Membrane Science, vol. 382, pp. 212-221, 2011. A. Reuvers and C. Smolders, Formation of membranes by means of immersion precipitation: Part II. the mechanism of formation of membranes prepared from the system cellulose acetate-acetone-water, Journal of membrane science, vol. 34, pp. 67-86, 1987. D. W. Wallace, C. Staudt-Bickel, and W. J. Koros, Efficient development of effective hollow fiber membranes for gas separations from novel polymers, Journal of membrane science, vol. 278, pp. 92-104, 2006. R. Boom, T. Van den Boomgaard, J. Van den Berg, and C. Smolders, Linearized cloudpoint curve correlation for ternary systems consisting of one polymer, one solvent and one non-solvent, Polymer, vol. 34, pp. 2348-2356, 1993. W. Koros and G. Fleming, Membrane-based gas separation, Journal of Membrane Science, vol. 83, pp. 1-80, 1993. C. Mahieux and K. Reifsnider, Property modeling across transition temperatures in polymers: a robust stiffness–temperature model, Polymer, vol. 42, pp. 3281-3291, 2001. R. F. Landel and L. E. Nielsen, Mechanical properties of polymers and composites: CRC Press, 1993. D. Ortiz-Acosta and C. Densmore, Sylgard® Cure Inhibition Characterization, Los Alamos National Laboratory (LANL)2012. S. Hussain, PhD. Thesis, Mixed matrix dual layer hollow fiber membranes for natural gas separation. , Georgia Institute of Technology, 2006.
Highlights 1. A new membrane material for sub-ambient temperature gas separation 2. Efficient carbon dioxide capture from flue gas 3. High membrane permeance combined with high selectivity
16
PII: DOI: Reference:
S0376-7388(16)30155-7 http://dx.doi.org/10.1016/j.memsci.2016.03.027 MEMSCI14365
To appear in: Journal of Membrane Science Received date: 11 November 2015 Revised date: 4 March 2016 Accepted date: 13 March 2016 Cite this article as: Lu Liu, Wulin Qiu, Edgar S. Sanders, Canghai Ma and William J. Koros, Post-combustion carbon dioxide capture via 6FDA/BPDADAM hollow fiber membranes at sub-ambient temperatures, Journal of Membrane Science, http://dx.doi.org/10.1016/j.memsci.2016.03.027 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.
Post-combustion carbon dioxide capture via 6FDA/BPDA-DAM hollow fiber membranes at sub-ambient temperatures Lu Liua, Wulin Qiu a, Edgar S. Sandersb, Canghai Mab, William J. Korosa* a School of Chemical & Biomolecular Engineering, Georgia Institute of Technology, 311 Ferst Drive, Atlanta, GA 30332-0100, United States b
AirLiquide, Delaware Research and Technology Center, 200 GBC Drive, Newark, DE 19702
*
Email: [email protected] Phone: 404-385-2845 Fax: 404-385-2683
Abstract Post-combustion carbon dioxide capture has the greatest near-term potential compared to other technological pathways to mitigate CO2 emissions. The hybrid membrane and cryogenic distillation process developed at American AirLiquide allows sub-ambient temperature (-50 °C to -20 °C) CO2 capture to be cost effective. In previous work, Matrimid® hollow fiber membranes with so-called “nodular” selective layers were successfully formed and showed a combination of high selectivity and high permeance at temperatures below -20 °C. Despite good performance, considering the low CO2 concentration and large volume of flue gas to be treated, the permeance of the “nodular” Matrimid® fibers still required improvement. In this work, a new membrane material with higher fractional free volume than Matrimid®, 6FDA/BPDA-DAM (1:1), was investigated for sub-ambient temperature (SAT) CO2 capture. Defect free 6FDA/BPDA-DAM hollow fiber membranes were successfully formed; however, the bare fibers lost selectivity at temperatures below 0 °C, possibly due to the development of minor defects in the thin selective layer. This problem was corrected by post-treatment with polydimethyl siloxane (PDMS). The PDMS post-treated 6FDA/BPDA-DAM hollow fiber membranes displayed much higher permeance than the Matrimid® fibers with a high selectivity. The high selectivity is presumably due to additional polymer backbone rigidity at SAT conditions combined with possible local chain orientation; and the high permeance is believed to be a result of hindered packing of the polymer backbone. Keywords: Carbon dioxide capture; CO2/N2 separation; Hollow fiber membrane; Sub-ambient temperature; 6FDA/BPDA-DAM 1. Introduction There is growing concern that human-related carbon dioxide emissions are contributing to a warming world [1-3]. Over the past century, climate scientists observed a dramatic increase in the atmospheric carbon dioxide concentration, from a pre-industrial value of about 280 ppm to 394 ppm in 2012 [4]. If no carbon dioxide emission mitigation measure is carried out, this value is predicted to increase to 540-970 ppm by 2100 [5]. Among the many human activities that produce carbon dioxide, the combustion of fossil fuels (coal, oil and natural gas) is the largest source of CO2 emissions. In 2011, about 82 % of the global total primary energy supply was 1
generated by the combustion of fossil fuels [4]. Compared to oil and gas, coal provides the lowest cost method of producing electricity and is relatively abundant. In the last ten years, the emissions of CO2 from coal combustion increased significantly, which surpassed oil to become the largest source of CO2 emissions [4]. A recent study at MIT concluded that coal consumption will continue to increase into the foreseeable future [6]; therefore, it is important to control CO2 emissions from coal combustion. In previous work, we investigated post-combustion carbon dioxide capture via Matrimid® hollow fiber membranes at SAT conditions [7]. Asymmetric Matrimid® fibers with so-called “nodular” selectivity layers were post-treated with polydimethylsiloxane (PDMS) to provide a combination of high permeance and high selectivity at temperatures between -50 °C and -20 °C. Although cooling the flue gas to temperatures below -20 °C requires considerable energy input, the hybrid membrane and cryogenic distillation process developed at American AirLiquide makes the capture of carbon dioxide at such low temperatures cost effective. A detailed description of the process can be found in previous work [8, 9]. The coupling of the unique high performing membranes with cryogenic processing technology ensures efficient capture of carbon dioxide. The low CO2 concentration and large volume of flue gas to be treated make membrane productivity very important, so higher CO2 permeance is an attractive target of this work. In this work, a new membrane material, 6FDA/BPDA-DAM (1:1), was investigated for SAT carbon dioxide capture. The full name of 6FDA/BPDA-DAM is 4,4’(hexafluoroisopropylidene) diphthalic anhydride (6FDA)/ 3,3’-4,4’-biphenyl tetracarboxylic acid dianhydride (BPDA) - 2,4,6-trimethyl-1,3-phenylene diamine (DAM). The high free volume, rigid 6FDA/BPDA-DAM polyimide has been well characterized by previous researchers in terms of its structure and ambient temperature intrinsic gas separation properties [10-12]. Moreover, 6FDA/BPDA-DAM-derived carbon molecular sieve (CMS) hollow fiber membranes show attractive gas separation properties [13-16] versus less rigid polymer precursors (e.g. Matrimid®). However, so far, there is no report on the gas separation properties of the 6FDA/BPDA-DAM fibers at sub-ambient temperatures. The high fractional free volume of 6FDA/BPDA-DAM contributes to higher membrane permeability than Matrimid® at room and high temperatures [12-14], and the same trend was expected at SAT conditions. In the current work, 6FDA/BPDA-DAM (1:1) was made into a hollow fiber configuration, and its gas separation properties at SAT conditions were investigated and compared with the high performing “nodular” Matrimid® hollow fiber membranes. 2. Experimental 2.1. Materials The asymmetric hollow fiber membranes studied in the current work was formed from the polyimide 6FDA/BPDA-DAM (1:1). It was synthesized via a two-step polycondensation reaction described in other work [12]. The first step involves the reaction of stoichiometric amounts of the dianhydrides 6FDA and BPDA in 1-methyl-2-pyrrolidinone (NMP) at temperatures around 5 °C to produce a high molecular weight polyamic acid; and the second step is an imidization reaction which closes the imide ring via dehydration. The 6FDA/BPDA-DAM (1:1) polymer was dried in a vacuum oven at 100 °C overnight before use. The structure of its 2
repeating unit is shown in Fig.1 and its physical properties are shown in Table 1. The pure gas transport properties of 6FDA/BPDA-DAM dense film membranes at 35 °C were studied by previous researchers and are shown in Table 2.
Fig.1. The repeating unit of 6FDA/BPDA-DAM. The ratio of X to Y is 1:1.
Table 1 Physical properties of 6FDA/BPDA-DAM (1:1) and Matrimid® [11]. Fractional Free Polymer Density (g/cm3) Tg (°C) Volume (FFV) Matrimid® 1.25 0.110 305 6FDA/BPDA-DAM (1:1)
1.32
0.145
424
Table 2 Pure gas transport properties of 6FDA/BPDA-DAM (1:1) dense films at feed pressures < 100 psia and 35 °C [11, 17] Permeability (Barrer) Material
6FDA/BPDA-DAM
1 Barrer 11010
Selectivity
O2
CO2 C2H4 C3H6
O2 N2
CO2 N2
CO2 CH 4
C2 H 4 C2 H 6
C3 H 6 C3 H 8
64
309
4.1
19.8
22
3.3
13
46
11.8
cm3 ( STP) cm cm2 s cmHg
2.2. Formation of asymmetric hollow fiber membranes The asymmetric hollow fiber membranes used in this work was formed via the wellknown “dry-jet/wet-quench” spinning technique. Detailed description of this process can be found in our earlier papers - based on simultaneous co-extrusion of the dope and bore fluid through a spinneret [7, 18-20]. Once extruded, the nascent fiber was drawn in an air gap, where solvent and/or non-solvent evaporation occurred at the outermost portion of the fiber to form a nascent ultra-thin selective layer. The nascent fiber then entered a water quench bath where phase separation occurred and a solid fiber was formed. The solid fibers were collected and immersed in deionized (DI) water for 4 days, with the water being changed daily. This promoted the removal of most of the trace residual solvent and non-solvent from the fibers. The fibers were 3
then sequentially soaked in low surface tension methanol and hexane for 1 hour, with the solvents being changed every 20 minutes. After solvent exchange, the fibers were air-dried in a fume hood for 1 hour and dried under vacuum at 75 °C for 3 hours. 2.3. Permeation Both pure and mixed gas permeation experiments in the current work were carried out in a constant pressure system, and a countercurrent flow configuration was used. For pure gas permeation testing, feed was introduced at the bore side of the hollow fiber membranes, and the permeate was collected at the shell side. The permeate flow rate was measured via a bubble flowmeter. A typical hollow fiber module in this work contains 2 fibers and has an effective length of around 20 cm. For mixed gas permeation, a bore side feed flowrate was maintained to prevent concentration polarization and maintain an essentially constant feed. In this work, the majority of the mixed gas permeation testing was carried out at temperatures from -50 °C to 35 °C. A BTZ475 benchtop temperature chamber (ESPEC North America, INC., Hudsonville, MI), which accurately controlled temperature within the range of -70 °C to 180 °C with ± 0.5 °C fluctuation, was used to contain the modules. The permeate flow rate was measured by a bubble flow meter and the permeate composition was analyzed by a gas chromatograph (Bruker 430-GC, Agilent Technologies, CA). The retentate flow rate was controlled by a needle valve and was measured by a digital mass flowmeter (FMA Series, Omega Engineering, Inc.) to provide low stage cuts, typically below 1%. 2.4. Pressure decay sorption The pressure decay method was used in the current work to characterize gas sorption by the polymer. The design considerations of a pressure-decay apparatus has been described earlier [21], and some modifications were made to construct the pressure-decay apparatus for sorption testing at cold conditions. The same BTZ-475 benchtop temperature chamber (ESPEC North America, INC., Hudsonville, MI) as used in the permeation system was selected to contain the pressure decay apparatus. The apparatus consists of two main compartments: the reservoir and the sample cell. The polymers were placed in the sample cell, and pressure transducers were used to monitor the pressures in each compartment. All the valves within in the temperature chamber were pneumatically-actuated, and were controlled by National Instruments LabVIEW 8.6 code. The equilibrium sorbed concentration of the gas was calculated through a standard mole balance. 2.5. Scanning Electron Microscopy (SEM) The morphology of the asymmetric hollow fiber membranes was examined by scanning electron microscopy (SEM). A high resolution SEM (LEO 1530) was used in this work. The LEO 1530 was equipped with a thermally assisted field emission gun, and an 8 kV operation voltage was found to display a clear image. In the preparation of samples, dried fibers were soaked in hexane for ~ 1 minute and then quickly transferred into liquid nitrogen. The fibers were shear fractured in liquid nitrogen and then mounted vertically on sample holders. Prior to examination, samples were sputter coated with gold for 30 seconds. 4
3. Results and discussion 3.1. Formation of 6FDA/BPDA-DAM hollow fiber membranes The formation of 6FDA-based hollow fiber membranes has been studied by several researchers in the past. Wallace [22] and Chen [23] successfully formed defect free asymmetric hollow fiber membranes from 6FDA-DAM/DABA (4:1) and 6FDA-DAM/DABA (3:2), respectively. Weinberg [24] studied the formation of 6FDA/BPDA-DAM hollow fiber membranes using dimethyl sulfoxide (DMSO) as one of the solvents, but no defect-free selective layer formation was reported. Recently, Xu [17, 25] successfully formed defect free 6FDA-DAM and 6FDA/BPDA-DAM asymmetric hollow fiber membranes. In this work, 6FDA/BPDA-DAM (1:1) with a somewhat lower molecular weight (123 kDa) than Xu’s polymer (160 kDa) was used as the membrane material; however, this had negligible impact on spinning. This polymer has a polydispersity index of 2.0 (Table 3). Table 3 Molecular weight and polydispersity index of 6FDA/BPDA-DAM (1:1) used in the current work Polymer
Mw (kDa)
PDI (Mw/Mn)
6FDA/BPDA-DAM (1:1)
123
2.0
Compared to the spinning of Matrimid® hollow fiber membranes, 6FDA/BPDA-DAM requires more non-solvent for phase separation at equivalent polymer concentration. Reflected on the ternary phase diagram, the binodal line for 6FDA/BPDA-DAM is closer to the nonsolvent vertex (Figure 2). Slower phase separation tends to occur for 6F-based polymers compared to Matrimid®, so lithium nitrate (LiNO3) was introduced to the dope as an additional non-solvent and pore forming component. Lithium nitrate complexes with the solvent NMP molecules in the dope and the complex dissociates in the water quench bath and thus accelerates phase separation [17]. A 6.5 wt. % lithium nitrate level was found in our earlier work to be an appropriate concentration [17, 26, 27], and the same concentration was used for the current work. N-methyl-2-pyrrolidone (NMP) and tetrahydrofuran (THF) were used as the solvents, and ethanol (EtOH) was selected as the non-solvent. A dope’s viscosity and ability to phase separate are two critical parameters for the formation of desirable hollow fiber membranes, which are determined by the initial dope composition. Adequate viscosity allows the dope to tolerate the rigorous extensional stresses during fiber spinning, and rapid phase separation promotes the formation of open pores in the substructure [28]. A dope’s initial composition should be close to the binodal line (the boundary between the one-phase and two-phase regions) on the ternary phase diagram. The binodal line in this work was determined via the “cloud point” technique [29, 30]. This technique involves the preparation of a series of dope samples with different compositions. In the current work, three polymer concentrations were selected: 16 wt.%, 20 wt.% and 25 wt.%. At each polymer concentration, dope samples with increasing non-solvent contents (and accordingly decreasing solvent contents) were prepared and well mixed. The dope samples with a composition sitting in 5
the one-phase region were transparent, whereas those with a composition sitting in the two-phase region appeared cloudy. The composition on the phase boundary was the “cloud point”, as is shown by the open circles in Fig. 2. All the cloud points were connected to form an approximate binodal line (the solid line in Fig. 2). The considerable difference in the binodal for this 6FDAbased polymer versus the simpler Matrimid® polymer (the dashed line in Fig. 2) is obvious. The solid red point on the ternary phase diagram is the initial dope composition used in the current work, and the detailed composition is listed in Table 4.
Fig.2. Initial dope composition and binodal line for the 6FDA/BPDA-DAM (1:1) used in the current work. The solid line is the binodal for 6FDA/BPDA-DAM, and the dashed line is the binodal for Matrimid®[17]. Open circles represent cloud points for 6FDA/BPDA-DAM in the current work, and the solid red point represents the initial dope composition. After two spinning attempts, defect free 6FDA/BPDA-DAM hollow fiber membranes were successfully formed. The spinning conditions for the successful attempt are listed in Table 5. A high spinneret temperature (70 °C) was used to promote the evaporation of volatile solvents and non-solvents in the air gap, and a high quench temperature (50 °C) was used to accelerate phase separation. The SEM images of the resultant fiber’s cross section and selective layer are shown in Fig. 3.
6
Table 4 Initial dope composition for the spinning of 6FDA/BPDA-DAM (123 kDa) hollow fiber membranes Component
wt.%
6FDA/BPDA-DAM
20
NMP
47.5
THF
10
Ethanol
16
LiNO3
6.5
Table 5 Spinning conditions for the formation of defect free 6FDA/BPDA-DAM hollow fiber membranes Spinning parameter
Value
Dope flow rate
180 ml/h
Bore flow rate
60 ml/h
Bore fluid composition
85%/15% NMP/H2O
Take-up rate
30 m/min
Quench temperature
49 °C
Spinneret temperature
70 °C
Air gap height Room temperature Relative humidity
10, 15, 20 cm 25 °C 33-36 %
7
Fig.3. SEM images for the cross-section and selective layer of the defect free 6FDA/BPDADAM hollow fiber membranes. Skin integrity of the defect free 6FDA/BPDA-DAM hollow fiber membranes was confirmed by permeating pure O2, N2 and CO2 at 35 °C. Feed pressure for O2 and N2 was 100 psig; and for CO2, 50 psig. The O2/N2 selectivity was 3.93 ± 0.11 (Table 6), which was more than 90 % of the 6FDA/BPDA-DAM dense film value, suggesting the hollow fiber membranes were essentially defect free. (The O2/N2 selectivity of 6FDA/BPDA-DAM dense film at 35 °C is 4.1 [10]). An asymmetric hollow fiber membrane is conventionally considered to be defect free if its selectivity for a given gas pair is higher than 90 % of the polymer’s intrinsic dense film value [20]. Table 6 Pure gas permeation results for defect free 6FDA/BPDA-DAM hollow fiber membranes at 35 °C. Feed pressure for O2 and N2 was 100 psig; for CO2, 50 psig. The feed was introduced at the bore side.
PO2 l
(GPU)
82.8 ± 4
PN 2 l
(GPU)
21.1 ± 0.5
PCO2 l
(GPU)
498 ± 8
O2 / N2
CO2 / N2
3.93 ± 0.11
23.7 ± 0.1
3.2. Mixed gas permeation To investigate the separation properties of defect free 6FDA/BPDA-DAM hollow fiber membranes at cold conditions, mixed gas permeation measurement was carried out. A mixture of 20 mol% CO2/80 mol% N2 was introduced at the bore side of the fibers at 100 psig. In the permeation calculation, fugacity was used instead of partial pressure to account for the nonidealities of gases, and pressure drop along the fibers were also considered. Starting from 35 °C, the temperature was reduced step by step, with the feed continuously purging the module. As 8
70
3500
60
3000
CO2 Permeance (GPU)
CO2/N2 Selectivity
temperature was decreased from 35 °C to 0 °C, an increase in selectivity and decrease in permeance was observed (Fig. 4), as was expected due to the activated nature of permeation. The higher CO2/N2 selectivity at lower temperature was attributed to the lower permeation activation energy for CO2 than for N2 [18, 31]. However, as temperature was decreased further, the membrane lost selectivity and a dramatic increase in permeate flow rate was observed. This experiment was repeated 4 times, and the same results were obtained. At first, it was suspected that water might be the cause for the sudden loss of selectivity. Crystallization of residue water in the fiber backbone might lead to change of overall volume, which might subsequently lead to rupture of fiber. However, the thermogravimetric analysis (TGA) results confirmed that there was minimal amount of residue water present in the fiber, as is shown in Fig. 5. No water peak was observed. The weight loss between 80 °C and 120 °C is less than 0.5%, indicating the amount of water in the 6FDA/BPDA-DAM fibers is negligible. As the permeation temperature was brought back from -40 °C to 35 °C, the membrane regained some selectivity. However, the new selectivity was slightly lower than the original value (~ 12 % lower), and a higher membrane permeance was observed (~ 28 % higher). This unusual phenomenon suggested the thin selective layer of the membranes might have developed some defects at temperatures below 0 °C, possibly due to stresses on the increasingly rigid polymer chains as temperature decreased [32, 33]. Similar phenomena were observed for defect free Matrimid® hollow fiber membranes, which started to lose selectivity at temperatures below 20 °C.
50 40 30 20 10 0 -50 -40 -30 -20 -10 0
10 20 30 40
2500 2000 1500 1000 500 0 -50 -40 -30 -20 -10 0
10 20 30 40 o
o
Temperature ( C)
Temperature ( C)
(a) (b) Fig.4. Mixed gas permeation results of defect free 6FDA/BPDA-DAM hollow fiber membranes at different temperatures. The feed was a mixture of 20 mol% CO2/80 mol% N2, and was introduced at the bore side at 100 psig.
9
80 70
0.6
6FDA/BPDA-DAM(1:1) fiber TGA heating at 10 K/min in N2
0.4
60 0.2 50 40
o
Residual weight (wt%)
0.8
90
Derivative weight (% / C)
1.0
100
0.0 100 200 300 400 500 600 700 800 900 o
Temperature C
Fig.5. TGA results of 6FDA/BPDA-DAM (1:1) fibers. Sample was heated directly from 30 °C to 900 °C in a N2 atmosphere. The heating rate was 10 K/min.
3.3. Performance of PDMS post-treated 6FDA/BPDA-DAM membranes PDMS post-treatment is known to be effective in plugging pinhole defects in the membrane’s selective layers and thus enhance membrane performance [7, 18]. We wondered whether such a treatment could help prevent the 6FDA/BPDA-DAM hollow fiber membranes from losing selectivity at cold conditions. In this work, a 2 wt.% Sylgard® 184 (a two-part, silicon elastomer kit [34]) was dissolved in heptane to prepare the PDMS solution. To achieve a desirable PDMS molecular weight, the solution was heated at 100 °C in an oil bath for 5 hours with stirring until a slight increase in viscosity of the solution was observed. This increase in viscosity reflected chain growth and incipient network formation of PDMS chains, which were effective in caulking minor defects in the membranes [35]. Fibers in a module were shell-side soaked in this mixture for 30 minutes, after which the solution was drained. Then the module was dried in a fume hood for 24 hours at room temperature, followed by curing at 78 °C for 2 hours under vacuum. The PDMS post-treated 6FDA/BPDA-DAM hollow fiber membranes were tested at subambient temperatures. The same testing conditions as described in Section 3.2 were used. The 20 mol% CO2/80 mol% N2 was introduced at the bore side of the fibers at 100 psig. Interestingly, the PDMS post-treated 6FDA/BPDA-DAM fibers didn’t lose selectivity when the temperature was dropped below 0 °C, as is shown in Fig. 6. The CO2/N2 selectivity kept increasing with the decrease of temperature over a range of -50 °C to 35 °C (Fig. 6a). At -50 °C, the CO2/N2 selectivity was 214 ± 19, more than 10 times higher than the selectivity at 35 °C (20.9 ± 0.6). The CO2 permeance decreased slightly at first with decreasing temperature, and then the trend was reversed (Fig. 6b). The subsequent increase in CO2 permeance was believed to largely
10
reflect the stronger increase in CO2 solubility versus N2 at lower temperatures. This factor mitigates CO2 permeance reduction, despite decreased diffusivity.
CO2/N2 Selectivity
200 150 100 50 0 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50
2500
CO2 Permeance (GPU)
PDMS post-treated Un-treated
250
PDMS post-treated Un-treated
2000 1500 1000 500
0 -60 -50 -40 -30 -20 -10 0 10 20 30 40 o
o
Temperature ( C)
Temperature ( C)
.
(a) (b) Fig.6. Comparison of the performance of PDMS post-treated and un-treated defect free 6FDA/BPDA-DAM hollow fiber membranes at different temperatures. The feed was a mixture of 20 mol% CO2/80 mol% N2, and was introduced at the bore side at 100 psig.
The PDMS post-treatment was found to decrease the rigidity of 6FDA/BPDA-DAM (1:1) fibers. Fig. 7 shows the results of dynamic mechanical analysis (DMA) of un-treated and PDMS post-treated 6FDA/BPDA-DAM fibers. The PDMS post-treated fibers showed lower storage modulus and lower stiffness compared to the un-treated sample, indicating better mechanical strength. The PDMS post-treatment was found earlier to prevent defect free Matrimid® hollow fiber membranes from losing selectivity at sub-ambient temperatures as well. To sum up, in order for 6FDA/BPDA-DAM and Matrimid® hollow fiber membranes to achieve satisfying performance at SAT conditions, PDMS post-treatment is necessary. 700
550 500 450 400 350
Un-treated fiber PDMS treated fiber
1250
600 Stiffness (N/m)
Storage Modulus (MPa)
1300
Un-treated fiber PDMS treated fiber
650
1200 1150 1100 1050
40
60
80
100
120
140
40
o
60
80
100
120
140
o
Temperature ( C)
Temperature ( C)
Fig.7. DMA results of un-treated and PDMS post-treated 6FDA/BPDA-DAM (1:1) fibers. 11
3.4. Comparison of 6FDA/BPDA-DAM and Matrimid® hollow fiber membranes The performance of the PDMS post-treated 6FDA/BPDA-DAM hollow fiber membranes were also compared with the high performing “nodular” Matrimid® fibers developed in our previous work [7] as well as the defect free Matrimid® fibers. As discussed above, PDMS posttreatment is necessary for 6FDA/BPDA-DAM and Matrimid® fibers to achieve satisfying performance. Therefore, all fibers compared here were post-treated with PDMS. It was interesting to discover that the 6FDA/BPDA-DAM fibers showed much higher CO2 permeance than the Matrimid® fibers at SAT conditions; and at the same time, a high CO2/N2 selectivity was maintained (Fig. 8).
CO2/N2 Selectivity
250
Defect free 6F ® Nodular Matrimid ® Defect free Matrimid
200 150 100 50 0 -60 -50 -40 -30 -20 -10 0 10 20 30 40
700
CO2 Permeance (GPU)
300
o
600 500
Defect free 6F ® Nodular Matrimid ® Defect free Matrimid
400 300 200 100 0 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 o
Temperature ( C)
Temperature ( C)
(a) (b) Fig.8. Comparison of the performance of PDMS post-treated defect free 6FDA/BPDA-DAM and Matrimid® hollow fiber membranes at different temperatures. All fibers compared here are posttreated with PDMS. The feed was a mixture of 20 mol% CO2/80 mol% N2, and was introduced at the bore side at 100 psig.
The high CO2/N2 selectivity for 6FDA/BPDA-DAM fibers at SAT conditions was presumably due to additional polymer backbone rigidity coupled with possible local chain orientation; and the high CO2 permeance was believed to be due to higher CO2 solubility and higher CO2 diffusivity in the polymer compared to Matrimid®. The presence of bulky -CF3 groups in the rigid 6FDA/BPDA-DAM backbones hinders the packing of the polymer chains, leading to a larger Langmuir sorption capacity ( CH' ) than Matrimid®. The sorption of CO2 in 6FDA/BPDA-DAM at room and sub-ambient temperatures was carried out in the current work. The gas uptake was calculated by mole balance, and the non-ideality of CO2 was taken into account. Appropriate compressibility factors were used for the different operating temperatures, and fugacity was used instead of partial pressure. The CO2 sorption isotherms for 6FDA/BPDADAM at 35 °C and -40 °C are compared with the isotherms for Matrimid® (Fig. 9). At each temperature, a new sample was used to avoid the influence of polymer matrix swelling and conditioning due to the sorption of highly condensable CO2 at high partial pressures on measured gas uptake. For each set of data, two measurements were made, and a standard deviation of less 12
than 10% was observed. The data shown in Fig. 9 is the average of the two measurements. The discrete points are experimental data, and the solid lines are the fit of concentration versus fugacity data to the dual-mode sorption model. As is shown in Fig. 9, the gas uptake in 6FDA/BPDA-DAM was much higher than in Matrimid®, confirming the higher expected CO2 sorption in 6FDA/BPDA-DAM. The dual mode parameters derived from the sorption isotherms are listed in Table 7. The larger Langmuir sorption capacity ( CH' ) for CO2 in 6FDA/BPDA-
70
6FDA/BPDA-DAM ® Matrimid
60
CO2 Concentration (CCSTP/CC)
CO2 Concentration (CCSTP/CC)
DAM, a reflection of the amount of unrelaxed free volume in the samples being tested, is consistent with its higher solubility in the polymer. Since the selective layer thickness of the PDMS post-treated membranes cannot be determined reliably, actual diffusion coefficients cannot be determined. Nevertheless, the high free volume of 6FDA/BPDA-DAM may contribute to a larger number of transient gaps for CO2 molecules to jump through, resulting in higher CO2 diffusivity.
o
50
35 C
40 30 20 10 0
0
20
40
60
80
100
120
140
450
6FDA/BPDA-DAM ® Matrimid
400 350
o
- 40 C
300 250 200 150 100 50 0
0
10
20
Fugacity (psia)
30
40
50
60
70
Fugacity (psia)
Fig.9. The CO2 sorption isotherms for 6FDA/BPDA-DAM (1:1) at 35 °C and -40 °C as compared to Matrimid®. The discrete points are experimental data and the solid lines are the fit of concentration versus fugacity data to the dual-mode sorption model. The 6FDA/BPDA-DAM (1:1) samples for the sorption measurement has a molecular weight of 123 kDa.
Table 7 Dual-mode sorption parameters for CO2 in 6FDA/BPDA-DAM (1:1) and Matrimid® CH' (
T (°C)
35 -40
cm3 STP cm3 poly
)
6FDA/BPDA -DAM
Matrimid®
35.9 ± 1.2 112 ± 3
22.5 ± 2.5 47.4 ± 1.0
1 b psia 6FDA/BPDA Matrimid® -DAM
0.06 ± 0.01 0.86 ± 0.14
13
0.07 ± 0.01 0.82 ± 0.08
kD (
cm3 STP
cm3 poly psia
)
6FDA/BPDADAM
Matrimid®
0.19 ± 0.01 3.63 ± 0.06
0.14 ± 0.02 1.47 ± 0.03
4. Conclusions The rigid, packing inhibited 6FDA/BPDA-DAM (1:1) polyimide was investigated in this work for SAT conditions carbon dioxide capture. Defect free 6FDA/BPDA-DAM (1:1) hollow fiber membranes were successfully formed via the optimization of initial dope composition and spinning conditions. Mixed gas permeation experiments on the bare 6FDA/BPDA-DAM fibers showed a significant loss of selectivity and a dramatic increase in membrane permeance at temperatures below 0 °C. This was attributed to the development of minor defects in the membrane’s thin selective layer. However, it was interesting to discover that treating the 6FDA/BPDA-DAM fibers in a PDMS solution could correct this problem. The PDMS posttreated 6FDA/BPDA-DAM fibers showed increasing CO2/N2 selectivity as temperature was decreased from 35 °C to -50 °C. Compared to the high performing “nodular” Matrimid® hollow fiber membranes developed in a previous work, the 6FDA/BPDA-DAM fibers showed much higher permeance at SAT conditions; meanwhile, a high CO2/N2 selectivity was maintained. Considering the low CO2 concentration and large volume of flue gas to be treated, a higher membrane permeance will contribute to more efficient CO2 capture. The high permeance of 6FDA/BPDA-DAM hollow fiber membranes was attributed to the hindered packing of the polymer backbones, leading to higher CO2 solubility and diffusivity than the more densely packed Matrimid®.
Acknowledgement The authors would like to thank American AirLiquide for the funding of this project. References [1] I. P. O. C. Change, Climate change 2007: The physical science basis, Agenda, vol. 6, p. 333, 2007. [2] H. Yang, Z. Xu, M. Fan, R. Gupta, R. B. Slimane, A. E. Bland, et al., Progress in carbon dioxide separation and capture: A review, Journal of Environmental Sciences, vol. 20, pp. 14-27, 2008. [3] D. Aaron and C. Tsouris, Separation of CO2 from flue gas: a review, Separation Science and Technology, vol. 40, pp. 321-348, 2005. [4] CO2 Emissions From Fuel Combustion Highlights 2013 Edition, International Energy Agency, 2013. [5] J. T. Houghton, Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden, X. Dai, et al., Climate change 2001: the scientific basis vol. 881: Cambridge University Press Cambridge, 2001. [6] The Future of Coal - Options for a Carbon-Constrained World, MIT Interdisciplinary Study, 2007. [7] L. Liu, E. S. Sanders, S. S. Kulkarni, D. J. Hasse, and W. J. Koros, Sub-ambient temperature flue gas carbon dioxide capture via Matrimid® hollow fiber membranes, Journal of Membrane Science, vol. 465, pp. 49-55, 2014.
14
[8]
[9] [10] [11] [12] [13] [14]
[15]
[16]
[17] [18]
[19] [20]
[21]
[22]
[23] [24] [25]
D. J. Hasse, S. S. Kulkarni, E. S. Sanders Jr, J.-P. Tranier, and P. Terrien, Method of obtaining carbon dioxide from a carbon dioxide-containing gas mixture, ed: Google Patents, 2013. D. Hasse, S. Kulkarni, E. Sanders, E. Corson, and J.-P. Tranier, CO2 capture by subambient membrane operation, Energy Procedia, vol. 37, pp. 993-1003, 2013. K. M. Steel, Ph.D. thesis, Carbon membranes for challenging gas separations, The University of Texas at Austin, 2000. M. Rungta, Ph.D. thesis, Carbon molecular sieve dense film membranes for ethylene/ethane separations, Georgia Institute of Technology, 2012. M. Kiyono, Ph.D. thesis, Carbon molecular sieve membranes for natural gas separations, Georgia Institute of Technology, 2010. K. M. Steel and W. J. Koros, An investigation of the effects of pyrolysis parameters on gas separation properties of carbon materials, Carbon, vol. 43, pp. 1843-1856, 2005. D. Q. Vu, W. J. Koros, and S. J. Miller, High pressure CO2/CH4 separation using carbon molecular sieve hollow fiber membranes, Industrial & engineering chemistry research, vol. 41, pp. 367-380, 2002. M. Kiyono, P. J. Williams, and W. J. Koros, Effect of polymer precursors on carbon molecular sieve structure and separation performance properties, Carbon, vol. 48, pp. 4432-4441, 2010. L. Xu, M. Rungta, M. K. Brayden, M. V. Martinez, B. A. Stears, G. A. Barbay, et al., Olefins-selective asymmetric carbon molecular sieve hollow fiber membranes for hybrid membrane-distillation processes for olefin/paraffin separations, Journal of Membrane Science, vol. 423, pp. 314-323, 2012. L. Xu, Ph.D. thesis, Carbon molecular sieve hollow fiber membranes for olefin/paraffin separations, Georgia Institute of Technology, 2012. L. Liu, E. S. Sanders, J. R. Johnson, O. Karvan, S. Kulkarni, D. J. Hasse, et al., Influence of membrane skin morphology on CO2/N2 separation at sub-ambient temperatures, Journal of Membrane Science, vol. 446, pp. 433-439, 2013. D. T. Clausi and W. J. Koros, Formation of defect-free polyimide hollow fiber membranes for gas separations, Journal of Membrane Science, vol. 167, pp. 79-89, 2000. S. B. Carruthers, G. L. Ramos, and W. J. Koros, Morphology of integral‐skin layers in hollow‐fiber gas‐separation membranes, Journal of applied polymer science, vol. 90, pp. 399-411, 2003. W. J. Koros and D. Paul, Design considerations for measurement of gas sorption in polymers by pressure decay, Journal of Polymer Science: Polymer Physics Edition, vol. 14, pp. 1903-1907, 1976. D. W. Wallace, Ph.D. thesis, Crosslinked hollow fiber membranes for natural gas purification and their manufacture from novel polymers, The University of Texas at Austin, 2004. C.-C. Chen, Ph.D. thesis, Thermally crosslinked polyimide hollow fiber membranes for natural gas purification, Georgia Institute of Technology, 2011. M. G. Weinberg, Polyimide gas separation membranes for carbon dioxide enrichment, ed: Google Patents, 1993. L. Xu, C. Zhang, M. Rungta, W. Qiu, J. Liu, and W. J. Koros, Formation of defect-free 6FDA-DAM asymmetric hollow fiber membranes for gas separations, Journal of Membrane Science, vol. 459, pp. 223-232, 2014. 15
[26] [27]
[28]
[29]
[30]
[31] [32] [33] [34] [35]
I. C. Omole, Ph.D. thesis, Crosslinked polyimide hollow fiber membranes for aggressive natural gas feed streams, Georgia Institute of Technology, 2008. C.-C. Chen, W. Qiu, S. J. Miller, and W. J. Koros, Plasticization-resistant hollow fiber membranes for CO2/CH4 separation based on a thermally crosslinkable polyimide, Journal of Membrane Science, vol. 382, pp. 212-221, 2011. A. Reuvers and C. Smolders, Formation of membranes by means of immersion precipitation: Part II. the mechanism of formation of membranes prepared from the system cellulose acetate-acetone-water, Journal of membrane science, vol. 34, pp. 67-86, 1987. D. W. Wallace, C. Staudt-Bickel, and W. J. Koros, Efficient development of effective hollow fiber membranes for gas separations from novel polymers, Journal of membrane science, vol. 278, pp. 92-104, 2006. R. Boom, T. Van den Boomgaard, J. Van den Berg, and C. Smolders, Linearized cloudpoint curve correlation for ternary systems consisting of one polymer, one solvent and one non-solvent, Polymer, vol. 34, pp. 2348-2356, 1993. W. Koros and G. Fleming, Membrane-based gas separation, Journal of Membrane Science, vol. 83, pp. 1-80, 1993. C. Mahieux and K. Reifsnider, Property modeling across transition temperatures in polymers: a robust stiffness–temperature model, Polymer, vol. 42, pp. 3281-3291, 2001. R. F. Landel and L. E. Nielsen, Mechanical properties of polymers and composites: CRC Press, 1993. D. Ortiz-Acosta and C. Densmore, Sylgard® Cure Inhibition Characterization, Los Alamos National Laboratory (LANL)2012. S. Hussain, PhD. Thesis, Mixed matrix dual layer hollow fiber membranes for natural gas separation. , Georgia Institute of Technology, 2006.
Highlights 1. A new membrane material for sub-ambient temperature gas separation 2. Efficient carbon dioxide capture from flue gas 3. High membrane permeance combined with high selectivity
16