Layer-by-layer assembly of carbide derived carbon-polyamide membrane for CO2 separation from natural gas

Energy 157 (2018) 188e199

Contents lists available at ScienceDirect

Energy journal homepage: www.elsevier.com/locate/energy

Layer-by-layer assembly of carbide derived carbon-polyamide membrane for CO2 separation from natural gas Abdelrahman Awad a, Isam H. Aljundi a, b, * a b

Chemical Engineering Department, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia Center of Excellence in Nanotechnology, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 3 March 2018 Accepted 20 May 2018 Available online 21 May 2018

Carbide-derived-carbon (CDC) have garnered increasing attention because it has shown potential for various applications including gas separation. In this article, we report for the first time the fabrication of CDC/polyamide composite membrane for CO2/CH4 separation. Different loadings of CDC were embedded in the polyamide layer to develop CDC/Polyamide mixed matrix membranes (MMMs) by the interfacial polymerization reaction of piperazine (PIP) and isophthaloyl chloride (IPC). The morphology and structural properties of the fabricated membranes, as well as the CDC nanoparticles, were examined by SEM, FT-IR, TGA, XRD, and N2 adsorption analysis. The characterization results confirmed the successful incorporation of the CDC nanoparticles into the rough polyamide layer. Gas permeation measurements of the fabricated CDC/PA membranes demonstrated an 88% and 49% enhancement in CO2 permeance and CO2/CH4 selectivity compared to the neat polyamide membrane. The CDC nanoparticles disrupted the polyamide matrix which resulted in a higher free volume to transport gases. In addition, MMMs were assembled layer-by-layer and their permeation tests revealed that building more than one selective layer on top of the polysulfone support increased the CO2/CH4 selectivity and decreased the CO2 permeance. MMMs with 10 selective layers showed the best separation performance with a CO2/CH4 selectivity of 24. © 2018 Elsevier Ltd. All rights reserved.

Keywords: CDC Polyamide Membrane Gas separation Natural gas

1. Introduction Current demographic trends suggest that the world population will grow by 24.6% and reach 9.1 billion by 2040 [1]. With this population growth, the demand for clean energy is certain to increase rapidly. The estimated consumption of natural gas (NG) in 2014 was 3393 billion cubic meters and this is projected to expand by 45% by 2040 [1]. Even though NG is considered as a green energy source compared with other fossil fuels, it contains impurities such as CO2 and H2S. Large amounts of CO2 in raw NG affect the selling price, damage pipelines and equipment, and lead to pipeline blockages during the transportation of liquefied natural gas (LNG) over long distances [2]. Therefore the amount of CO2 in NG must be reduced to the limit specified by transportation and distribution companies (<2%) [3]. Since the importance of clean NG is increasing, the search for a new effective, robust and efficient process for natural gas purification has been accentuated in recent

* Corresponding author. Chemical Engineering Department, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia. E-mail address: [email protected] (I.H. Aljundi). https://doi.org/10.1016/j.energy.2018.05.136 0360-5442/© 2018 Elsevier Ltd. All rights reserved.

research studies. CO2 capture methods using conventional absorption [4e8], adsorption [9], and cryogenic [10] processes have been shown to have serious disadvantages. On the other hand, membrane separation is simple, cheap, energy efficient and has a better environmental impact. These advantages make membrane gas separation a competitive technology for CO2 separation [11e14]. Over the past few decades, membrane based CO2 separation has been extensively investigated and developed using a large number of materials (polymeric, inorganic and hybrid) and different process schemes [12e15]. Polymeric membranes are cheap, mechanically stable and easy to prepare in different modules. Nevertheless, the tradeoff limitation and the issue of plasticization are a challenge for using polymeric membranes in industry. Although inorganic membranes overcome the trade-off limit in small scale production, much larger costs are incurred when manufacturing inorganic membranes. Moreover, improvements in reproducibility are needed for largescale production [14]. Recently, (MMMs) have been known to exceed the Robeson's tradeoff limits and to be highly processable and cheap compared to inorganic membranes [16e18]. It can be considered as a modified

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

organic membrane in which either micro or nanoparticles of inorganic fillers are embedded into an organic continuous matrix to form a heterogeneous membrane. Over the past few decades, tremendous advancements in MMMs for CO2 separation have been made by using different inorganic fillers such as zeolites [19e23], MOFs [24e28], silica [29e31], metal oxides [32,33], new nanomaterials [34e37], and carbon materials [38e42]. Obviously, MMMs provide a new route to benefit from the advantages of inorganic materials. However, the inorganic fillers are not completely utilized because some of the incorporated nanoparticles end up in the nonselective substrate layer. Also controlling the distribution of fillers and the filler/polymer compatibility are considered as constraints in MMMs [17]. In this regard, thin film composite (TFC) and thin film nanocomposite (TFN) membranes are types of membranes that have a selective layer prepared by interfacial polymerization (IP) reaction [43]. TFC/TFN membranes can provide a high level of separation performance by precisely controlling the distribution of inorganic nanoparticles so they can be deposited only within the skin of the top selective layer during the IP reaction [43]. The IP reaction is defined as a type of step-growth polymerization in which the polymerization reaction occurs at an interface between an aqueous solution containing one monomer (polyamine) and an organic solution containing a second monomer (polyacyl halides). The IP process produces polyamide (PA) membranes with the following advantages: the possibility of producing an ultrathin selective skin layer, the availability of a large number of monomers that can be used, and more importantly, their defect-free nature and ease of scale-up to the commercial modulus [44]. However, most, if not all, of the currently developed TFC/TFN membranes are focused on reverse osmosis [45e48] nano-filtration [49], and the pervaporation process [50]. The development of TCN/TFN membranes using IP for the application of gas separation still lags behind and needs further study. Nowadays, special carbon materials with a high surface area and nano-size pores are playing a significant role in the advancement of new technologies. With the advent of the nanotechnology revolution, carbon nanomaterials have garnered increasing attention from scientific researchers because they have shown potential for various applications including gas separation and storage. Carbidederived carbon is a tunable carbon material synthesized by extracting metals from metal carbide by leaching in supercritical water, halogenation at high temperatures, vacuum decomposition, and other methods. Back in the 1960s, carbide-derived carbon was produced as a byproduct from the metal chloride manufacturing process according to Eq. (1):

MeC þ x=2Cl2 /MeClx þ C

(1)

where Me is the metal [51]. According to the preparation method and conditions, amorphous or nano-crystalline structures can be synthesized [52]. Due to the layer-by-layer extraction of metal from the template metal carbide, precise control over the nanoparticles' properties (pore size and shape, surface chemistry and surface area) could be achieved by changing the operating conditions, composition and structure of the carbide precursor. Gogotsi and his co-workers [53] produced CDC with a pore size distribution comparable to the pore size of zeolite materials and narrower than certain carbon nanotubes (CNTs) and activated carbon. Ranjan et al. [54] synthesized TiC-CDC at 800  C and reported a methane gas uptake of 46 cm3/g for the prepared TiC-CDC nanoparticles which is higher than the gas uptake of both activated carbon (less than 35 cm3/g) and some MOFs (less than 10 cm3/g). The use of carbide-derived carbon as a membrane was

189

investigated by Hoffman et al. [55]. They fabricated an amorphous membrane structure and their permeation results demonstrated nitrogen permeability of 67 Barrer. In another work, Hoffman et al. [56], reported on CO2 and CH4 sorption analysis at 25  C of Ti3SC2CDC prepared at 600  C. The results demonstrated an adsorption of 125 and 55 cc/g for the CO2 and CH4; respectively. The differences in the gas sorption were attributed to the interactions between the carbon surface and gas molecules as well as the chemical affinity of gases on the CDC surface, rather than to the molecular sieving effect. Therefore, CDC offers many advantages in separation processes over polymers, zeolite, and porous carbon and provides higher gas sorption capacity. Despite the attractive characteristics and the enormous potential of CDC for gas separation the area has not been well-studied for this application neither as an inorganic membrane nor as a nanofiller to fabricate MMMs. Therefore, this work is an attempt to introduce new nanomaterials as a filler to delve deeper into TFN membranes for the application of CO2 separation. The objective of this research is to prepare, characterize, and evaluate the performance of CDC/PA multi-layer MMMs. The separation properties of the fabricated membranes were evaluated using the pure gases of CO2 and CH4 in a constant volume variable pressure apparatus. Furthermore, the effect of CDC loading and operating conditions (T, P) on the separation performance were studied. 2. Experimental 2.1. Materials All chemicals were purchased from Sigma-Aldrich except methanol and n-hexane which were obtained from Scharlau. Polysulfone pellets (Mw of 35000) were used to fabricate the membrane support. N,N-dimethylacetamide (DMAC) with a purity of 99.9% was used as a less volatile solvent, and inhibitor-free tetrahydrofuran (THF) with a purity of 99.9% was the primary volatile solvent, while absolute ethanol was selected as a non-solvent additive. PIP and IPC with a purity of 99% were used as the monomers for the IP reaction. Titanium Carbide Nano powder (TiC, 99 þ %) purchased from USA Research nanomaterials Inc., USA. CO2 and CH4 gases with a purity of 99.999% were purchased from Abdullah Hashim Company, KSA. 2.2. Synthesis of CDC nanoparticles Following the reported procedure by our group [57], titanium carbide in a quartz boat was inserted in a quartz tubular furnace and heated at a rate of 10  C/min to the desired temperature, while continuously purging with Argon to create an air/oxygen free closed system. When the furnace temperature reached the desired set point, pure chlorine gas was introduced at a flow rate of (10e13 cm3/min) for 3 h. After chlorination, the post treatment was carried out with hydrogen gas at the same final temperature for 1 h to remove the remaining chlorine from the CDC which enhances the surface area and micro-pore volume of the nanoparticles. Then, the furnace was purged with argon gas to cool it down to an ambient temperature. 2.3. Membranes preparation 2.3.1. Polysulfone support PSF support was fabricated by a dry/wet phase inversion technique. Prior to membrane preparation, the PSF polymer pellets were dried overnight at 100  C in a vacuum oven in order to completely remove the moisture from the polymer. Dry PSF pellets were dissolved in a mixture of DMACs and THF, then ethanol was

190

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

added and the solution was stirred for 24 h at 25  C using a magnetic stirrer. The polymeric solution (Table 1) was then degassed at room temperature for 24 h to remove air bubbles. After that, the solution was caste on a clean glass plate with a thickness of 200 mm using a casting knife. The membrane was left in air for 60 s at ambient conditions and subsequently immersed in a de-ionized (DI) water bath for 24 h. The prepared membranes were immersed in methanol for 2 h for solvent-exchange and treated with Polydimethylsiloxane (3% in hexane) to eliminate infinitesimal defects or pinholes in the membrane, and then finally dried in a vacuum oven at 100  C for 48 h. 2.3.2. Preparation of polyamide and CDCs/PA mixed matrix membranes (MMMs) The Polyamide membrane was prepared using an IP method in which a polymerization reaction occurs between piperazine (PIP in DI water) and isophthaloyl chloride (IPC in hexane). The previously prepared PSF support was saturated with PIP solution (2%w/v) for 10 min and then a rubber roller was used to remove the excess PIP solution. Subsequently, PSF support was immersed in an organic solution (0.2%w/v IPC) for 3 min and the excess unreacted IPC was removed using pure hexane. Finally, the membranes were dried at 80  C for 10 min and the fabricated polyamide membranes were kept in DI water. For the thin film nanocomposite membrane, CDCs nanoparticles were incorporated into the polyamide layer during the interfacial polymerization reaction by adding the required amount (as shown in Table 2) of nanoparticles to 100 mL of the organic solution (IPC), then the solution was sonicated for 15 min using a probe sonicator and 1 h using a bath sonicator. After that the IP reaction was conducted following the same procedure used for polyamide membranes. In order to synthesis a layer-by-layer membrane structure, the interfacial polymerization described above was repeated several times. After a complete cycle of the interfacial polymerization reaction the membrane is considered to have one deposition layer. Then the same procedure was repeated again to form the second layer and so on. 2.4. Characterization The surface morphology of the membranes and CDC nanoparticles was studied using a scanning electron microscope (FESEM, MIRA 3, Tescan). Membrane samples were coated with a 5 nm gold layer prior to SEM analysis using an Ion sputter coater (Q150R S, Quorum Technologies). Moreover, the fabricated membranes, as well as the CDC nanoparticles, were analyzed by a Fourier transformation infrared spectrometer (FT-IR, Nicolet 6700, Thermo Electron Corporation), X-ray diffraction (XRD, D8-Advance, Bruker, Table 1 Composition and amount of the dope solution.

and thermogravimetric analysis (TGA, STA 449 F3 Jupiter, Netzsch) in the temperature range of 30e800  C at a heating rate of 10  C/ min under a nitrogen flow of 100 mL/min. Furthermore, nitrogen adsorption analysis was conducted by an Autosorb iQ-MP-XR (Quantachrome, USA) and selected adsorption data was used to calculate the surface area of CDC while the total pore volume ðV t Þ was estimated at a relative pressure (P/Po) of 0.99. The pore size distribution of CDC was determined using Non-Local Density Functional Theory (NLDFT) from the N2 adsorption branch. The micropore volume ðV micro Þ was calculated using the t-plot method and the difference between the total pore and the micropore volumes is known as the mesopore volume ðV meso Þ. 2.5. Gas permeation measurements The permeance of the fabricated membranes was examined using pure CO2 and CH4 gases. The membranes were tested at different feed pressures from 1 to 5 bars and at temperatures ranging from 300 to 323 K. Gas permeation experiments were conducted using the well-known constant volume/variable pressure (time-lag) method [58e60]. As shown in Fig. 1, the gas permeation setup was built to work in two modes (constant volume/variable pressure and constant pressure/variable volume). The system contains three Bronkhost Coriolis mass flow controllers, a membrane module, a permeation volume, a vacuum pump that is connected to the permeate volume, and pressure transducers to detect the feed and permeate pressures. The process is controlled by software and the data is collected by LabVIEW. A membrane sample with an effective area of 4.91 cm2 was cut and fixed inside the membrane cell and both sides of the membrane module were evacuated to a low pressure of less than 1 mbar. The gas is then fed into the module at a constant pressure. To determine the gas permeation, the valve used for the evacuation was closed and the pressure change in the permeate side was monitored over time. The leak rate was measured at the start of each experiment to obtain an accurate permeation rate and the data reported here is the average of at least two independent measurements. The permeance (p) was calculated in Gas Permeation Units (1 GPU ¼ 106 cm3 (STP)/(s cm2 cm-Hg)) using Eq. (2) [61]:

p¼

273:15  106 Vd
 76 A T 760 P2  14:7

     dP1 dP1  dt ss dt leak

(2)

where Vd is the downstream volume (cm3), A is the effective membrane area (cm2), P2 is the upstream pressure (psi), and   dP1 is the rate of pressure change in the downstream chamber dt ss   dP1 is the leak rate in mmHg at the steady state in mmHg s , and s , and dt leak

Component

Amount (g)

Concentration (%)

PSF Pellets DMAC THF Ethanol

4.00 5.81 5.81 1.75

23.03 33.45 33.45 10.07

Table 2 Composition of the prepared TFN membranes. Membrane

CDC %

PA MMM1 MMM2 MMM3 MMM4 MMM5

0 0.0005 0.002 0.1 0.5 1

T is the cell temperature in K. The selectivity of gas i to j ðaij Þ was estimated by Eq. (3):

.

aij ¼ pi pj

(3)

3. Results and discussion 3.1. CDC characterization FE-SEM (Fig. 2a-b) was used to study the surface morphology of the untreated titanium carbide powder (TiC) and the CDC was prepared at 800  C. The precursor carbide and CDC particles look similar and exhibit preservation of particle shape with slight

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

191

Fig. 1. Flowsheet of the setup used in the permeation experiments.

(a)

(c)

(b)

Fig. 2. (a) FE-SEM image of untreated TiC powder (b) FE-SEM image of CDC, (c) TEM image of CDC.

900 TiC powder CDC

4

600

Pore size distribution CDC

3 dV(r) cc/g/nm

N2 Adsorbed Volume (cc/g) @STP

shrinkage and some etching signs. This conformal transformation and shape-conservation of carbide materials during the production of CDC were observed by other researchers [62]. TEM images (Fig. 2c) of the prepared CDC materials showed that the CDC was a mixture of amorphous and ordered graphitized carbon. In previous reports [57], it was observed that the degree of ordering was directly proportional to the chlorination temperature and that the reaction temperature has a pronounced effect on the crystallinity of CDC. Fig. 3 shows the nitrogen adsorption isotherms of untreated TiC powder as well as the prepared CDC. It was observed that untreated TiC powder has an insignificant adsorption capacity for the nitrogen gas indicating a nonporous structure with a low surface area. On the other hand, the prepared CDC showed a type I adsorption isotherm [63] with a considerable increase of adsorbed volume at a low relative pressure which indicates the microporous nature of CDC. Moreover, the steep increase in the adsorbed amount of nitrogen at a low relative pressure (P/Po) implies the existence of a narrow pore size distribution. Fig. 3 shows the pore size distribution of untreated TiC powder and the prepared CDC samples determined by the NLDFT with a slit shape pore model. Before chlorination, no porosity was observed for TiC powder. However, a porous structure was developed after chlorination with a narrow pore size distribution.

300

TiC powder

2

1

0 0

1

2 Half pore width (nm)

3

4

0 0

0.2

0.4

0.6

0.8

1

Relative pressure (P/Po) Fig. 3. Nitrogen adsorption isotherms of untreated TiC powder and the prepared CDC at 77 K (inset is the pore size distribution).

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

Table 3 presents different structural properties of the prepared CDC and untreated TiC powder. Remarkable enhancements were achieved for all of the structural properties of CDC compared to that of the pristine carbide powder. For example, the micrpore area of the prepared CDC increased 261 times, while the BET surface area increased 66 times over that of the original TiC powder. The observed propertires (high surface area, micro porosity, and narrow PSD) of the prepared CDC are of great importance for the application of gas separation. Fig. 4 represents the FT-IR analysis of the prepared CDC. It shows that the CDC nanoparticles have small peaks at 1600 cm1 and 1215 cm1 which are ascribed to the aromatic carbon bond [64] and to the CeN and CeO bonds [65]; respectively. The small peak at 3420 cm1 may be ascribed to the adsorption of water from the ambient air. Similar observations were reported in the literature for CDC nanoparticles prepared at different temperatures [65,66].

60

Transmittance (%)

192

45

30

15

0 0

500

1000

1500

2000

2500

3000

3500

4000

Wavelength (cm-1)

Fig. 4. FT-IR analysis of the prepared CDC.

3.2. Membrane characterization 3.2.1. Membrane morphology In order to study the surface morphology of the fabricated membranes, SEM analysis was carried out as shown in Fig. 5. The highly porous PSF surface (Fig. 5a) was observed for membranes prepared by the convertional wet phase inversion method. These membranes showed very low gas selectivity. Therefore, a convective air evaporation period (60 s) was applied to the membranes before coagulation in the DI water bath which resulted in an asymmetric PSF structure with a top dense layer (Fig. 5b). It is believed that this structure was formed as a result of the phase separation, along with the instantaneous de-mixing in a quaternary system (polymer, solvent, non-solvent, and an additive) because the DMAC solvent has a high mutual affinity for water. Moreover, the dense layer was built as a result of the drying period before coagulating the membranes in the water bath [67e70]. During the initial step of the dry inversion process, THF was removed from the surface of the dope solution because it has the lowest boiling point (65.48  C). The SEM images of PA and CDC/PA MMMs (Fig. 5c-h) confirmed the formation of thin composite polyamide layers on top of the PSF support membrane. They also show that the smooth PSF surface was totally covered by a rough and nodular structure of polyamide as a result of the reaction between PIP and IPC. In the SEM images, CDC nanoparticles were observed on the surface and distributed well in the PA layer up to a certain concentration (a loading of 0.5%). Some agglomeration of nanoparticles was observed at a relatively high concentration (loading of 1%) which may create defects in the membranes. 3.2.2. Fourier transformation infrared (FT-IR) FT-IR spectra of the entire polyamideepolysulfone composite film and the CDCs/polyamide thin nanocomposite layer were scanned in the range of 400 cm1 to 4000 cm1 to study the chemical composition of the membrane samples. The bands can be ascribed to the interfacially polymerized layer, as well as the PSF support skin because the depth of the beam penetration is thicker than the polyamide layer [71]. Fig. 6 shows the FT-IR spectra of bare polyamide and CDC/PA MMMs with a CDCs loading of 0.1% and 0.5%. The band at 1680 cm1 is ascribed to the amide I (C]O) stretch, whereas the broadband at 1590 cm1 is attributed to the

amide II (CeN) stretch. Moreover, the band of the amine group (NH stretch) was observed at a wavelength of 1500 cm1. Therefore, the existence of amide I and II in the FT-IR spectra confirmed the formation of the polyamide layer as a result of the IP reaction. The amide bands were observed for CDC/PA membranes and a new peak at 1640 cm1 appeared as a result of embedding carbidederived carbon material into the polyamide film and this peak is ascribed to the aromatic carbon bond. We observed that the characteristic FT-IR peak of CDC nanoparticles at (1600 cm1) appeared in CDC/PA MMMs with a small transition. Therefore, FT-IR confirmed the occurrence of IP, as well as the successful addition of CDC nanoparticles. 3.2.3. Thermogravimetric analysis (TGA) The thermal behavior of CDC, bare PA, and CDCs/PA MMMs was studied by TGA as shown in Fig. 7. Nearly 5e6 mg of samples were heated at a constant heating rate of 10  C/min ranging from room temperature to 800  C in an inert N2 atmosphere at a flow rate of ml . The CDC nanoparticles experienced major weight loss in the 100min temperature range of (51e90  C) and the weight at the end of the cycle was 76.45%. Thermogravimetric analysis of TFC/TFN membranes demonstrated that for all membranes there was no considerable weight loss up to a temperature of 480  C, which is comparable to the thermal stability of other prepared polyamides reported in the literature [72]. However, pure polyamide membranes started to lose weight significantly at a temperature of approximately 480  C as the weight reduced from 98% to 33.3% at a temperature of 616  C. Even though all of the membrane samples of bare polyamide and MMMs seem to exhibit a similar temperature range (480e620  C) for major weight loss, the onset decomposition temperature for the bare polyamide membrane was lower than the onset temperature for membranes with CDC. Generally speaking, inorganic fillers have good thermal strength in comparison to the polymeric material. Upon heating, CDC materials absorbed the heat, and consequently suppressed the rate of membrane decomposition which resulted in relatively higher decomposition temperatures [73]. Therefore, embedding CDC nanoparticles into a PA matrix enhanced the thermal strength of the membranes. The thermal stability could be evaluated more accurately in terms of the Td50% and Td60% values (the temperatures at which the tested

Table 3 Different structural properties of the prepared CDC compared to untreated TiC powder. Sample

BET Surface area (m2/g)

Amicropore (m2/g)

Aexternal (m2/g)

V t (m3/g)

V micro (m3/g)

V meso (m3/g)

Average pore size (nm)

TiC powder CDC

39 2589

7 2340

31 249

0.051 1.216

0.002 0.984

0.049 0.232

e 1.542

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

193

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

Fig. 5. Surface SEM images of (a) porous PSF, (b) dense PSF, (c) MMM0, (d) MMM1, (e) MMM2, (f) MMM3, (g) MMM4, (h) MMM5.

membrane loses 50, and 60 of its initial mass; respectively. The TGA data showed that the values of Td50% and Td60% for the bare PA occurred at a temperature of 534  C, and 546  C; respectively, while for the mixed matrix membranes Td50% and Td60% occurred in the range (543e546) C and (574e586) C; respectively. 3.2.4. X-ray diffraction (XRD) A Bruker X-ray diffractometer was used to take XRD spectra of bare PA and CDC/PA MMM with a wavelength of 1.5414 Å. The angle of diffraction (2Ѳ) was changed from 5 to 50 in order to observe the nature of the membrane structures. A scanning rate of 2 /min operating at 30 kV and 30 mA was used for the XRD spectra. Fig. 8 shows the microstructure of CDC, PA, and CDC/PA MMMs. It can be seen that the CDC has two peaks at 2Ѳ of 26 and 43 . The peak at 26 is attributed to (002) planes of graphite, while the peak at 43 corresponds to diffraction from the (101) planes of graphite [52]. These peaks implied the crystalline structure of the prepared th prepared CDC. Similar findings were reported in literature for CDC

prepared at 800  C [52,54]. The structure of CDC nanoparticles depends on the preparation conditions. Generally speaking, amorphous CDCs can be prepared at low temperatures while higher temperatures produces crystalline CDC [53]. As shown in the XRD spectra of the PA membrane, a broad peak centered around 19 revealed the amorphous structure [46]. While the very small peak around 21 indicates the semi-crystalline nature of the composite thin layer. The presence of crystalline regions can be attributed to the thin PA layer, while amorphous regions appeared as a result of the PSF structure [74]. The XRD patterns of the CDC/PA membranes with different loading were almost identical to that of the bare PA membrane. Therefore, the crystallinity of the pure PA membrane was conserved upon the addition of CDC nanoparticles. 3.3. Gas permeation measurement 3.3.1. Pure gas permeance and selectivity The gas separation performance of all fabricated membranes

194

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

250

PA MMM3 MMM4

Transmittance %

200

150

100

50

0

0

500

1000

1500

2000

2500

3000

Wavelength (cm-1) Fig. 6. FT-IR spectra of Polyamide and CDC/PA MMMs with the loading of 0.1% and 0.5%.

120

CDC PA MMM3 MMM4

100

wt%

80 60 40 20 0 0

200

400

600

800

T (°C) Fig. 7. TGA analysis for CDC, polyamide membrane and CDC/polyamide MMMs with different nanoparticle loadings.

Fig. 8. XRD Pattern of CDC nanoparticles, PA, and CDC/PA mixed matrix membranes.

was evaluated using pure CO2 and CH4 gases at a temperature of 300.15 K and a feed pressure of 5 bar as shown in Table 4. It is known that the amide and free amine groups in the PA membrane render the facilitated transport mechanism of CO2 and these functional groups are activated in the presence of moisture on the surface [75]. Our membrane samples were saturated by DI water before testing the gases to benefit from the facilitated transport mechanism of CO2. Therefore, we believe that the fast permeation of CO2 can be explained by both the facilitated transport and the solution diffusion mechanisms. The overall experimental results shown in Table 4 revealed that the CDC/PA membranes experienced higher gas permeance and selectivity in comparison with the reference bare PA membranes. The enhanced gas permeation and selectivity was primarily attributed to the addition of CDC nanoparticles which facilitated faster gas flow by disrupting the polymer chain matrix. As confirmed by the SEM images in Fig. 5, the CDC nanoparticles dispersed well in the PA layer. This resulted in higher gas permeance since the well-dispersed CDC could construct channels in the PA matrix to transport gas molecules more effectively. Furthermore, it can be observed that the ideal gas selectivity of the PSF membrane was enhanced by building the thin film PA layer, while the gas permeance decreased as a result of the increased mass transfer resistance. In addition, the CO2 gas permeation for all membrane samples was greater than that of CH4 which followed the order of kinetic (CO2 ¼ 0.330 nm, CH4 ¼ 0.380 nm). Since CO2 is a soluble gas, it is believed that it undergoes dissolution in the PA film layer which contains polar eNHCO functional groups as confirmed by FT-IR analysis in Fig. 6. Moreover, CO2 gas is expected to form hydrogen bonding interactions with amide groups of the PA membrane. On the other hand, the saturated non-polar CH4 gas molecules exhibited poor interaction with the membrane and hence showed much lower permeance. In addition, a reversible reaction between the amine groups and the CO2 molecules is expected to occur in wet PA membranes which produce complex and HCO3e components that could diffuse freely across the membranes [76]. However, because methane did not react with the PA film, their transport across the membrane is based on a simple solution-diffusion mechanism [75]. As a result of the high permeance of CO2 compared with CH4, relativity high selectivities of 13.33 and 19.92 was recorded for the fabricated TFC and TFN membranes, respectively. 3.3.2. Gas sorption measurements The adsorption measurements of CO2 and CH4 by the CDC nanoparticles at 293.15 K are shown in Fig. 9. Both adsorption curves can be explained by the type I isotherm [63,77]. It can be seen that for the whole range of pressures, the CO2 uptake is remarkably larger than CH4. Selective gas adsorption is generally achieved by (1) molecular sieving, (2) the thermodynamic equilibrium mechanism, (3) the diffusion effect, and (4) the quantum effect. Since the mean pore size of the CDC nanoparticle (1.542 nm) is significantly larger than the kinetic diameter of both CO2 (0.330 nm) and CH4 (0.380 nm), we believe that the preferential adsorption of CO2 on CDC nanoparticles is based on the thermodynamic equilibrium effect which is in good agreement with other findings in the literature [53]. The adsorption capacities of other fillers and other prepared CDCs reported in the literature are tabulated and compared with the nanoparticles used in this study as shown in Table 5. Fig. 10 shows the CO2 and CH4 adsorption isotherms for neat PA and CDC/PA MMM at 293.15 K. In both cases the fabricated PA and CDC-MMMs were found to adsorb CO2 more favorably than CH4. No significant difference was noticed between the amount adsorbed by PA and the CDC/PA membranes since the amount used from CDC to fabricate the membrane is very low. As mentioned earlier, the

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

195

Table 4 Gas separation performance of the fabricated PSF, PA, and MMM4 at 300.15 K and 5 bar. Membrane

CDC Loading %

PSF PA MMM4

Permeance (GPU) CO2 2.41 ± 0.16 2.16 ± 0.04 4.07 ± 0.43

0 0 0.5

25

100

MMM-CO2 PA-CO2 MMM-CH4 PA-CH4

CO2

Amount adsorbed (cc (STP)/g)

Amount adsorbed (cc (STP)/g)

CH4 80

60

40

20

Selectivity CO2/CH4 4.33 ± 0.77 13.33 ± 0.36 19.92 ± 3.63

CH4 0.556 ± 0.004 0.162 ± 0.001 0.204 ± 0.015

20

15

10

5

0

0 0

20

40

60

80

100

0

120

Pressure (kPa) Fig. 9. Adsorption isotherms of CO2 and CH4 gases on CDC- nanoparticles at 293.15 K.

adsorption mechanism is greatly affected by the interaction between the gas and the membrane surface. In addition, CO2 has a high affinity for amine groups in the membranes. As a result, the lower gas uptake for PA compared to CDC-MMM can be ascribed to the weaker NH bond that is observed in FT-IR analysis for MMMs (Fig. 6) which resulted in reduced NH functionality and lower CO2 uptake. When the adsorption isotherms are based on the equilibrium effect (adsorbent pore size is larger than the gases), the interaction between the nanoparticles and the gas molecules is vitally important and surface characteristics such as polarizability, permanent dipole moment, and quadrupole moment are dominant in determining the separation quality [78]. Because CO2 has a large quadrupole moment, higher polarizability, and can form H-bonding with the surface of the membranes, all of the prepared membranes exhibited a higher gas uptake for CO2 compared to CH4.

3.3.3. The effect of nanoparticle loading The effect of CDC loading onto the gas permeation properties was studied by fabricating membranes with different CDC loadings as shown earlier in Table 2. Table 6 shows the change in CO2 and CH4 gas permeance by varying CDC loading from 0.0005% to 1%. It is obvious that the permeation rate of both the CO2 and CH4 gas

20

40

60 Pressure (kPa)

80

100

120

Fig. 10. CO2 and CH4 gas adsorption in pure polyamide and mixed matrix membranes membrane at 293.15 K.

molecules was enhanced by increasing the amount of CDC in the PA layer. This can be attributed to the superior properties of the CDC in terms of high surface area and porosity which offered a larger surface and a higher volume to enable gases to diffuse through the membrane matrix. Generally speaking, the incorporation of carbon materials into the polymeric matrix alters the polymer chain packing which results in higher free volumes and improved gas diffusion. Even though both CO2 and CH4 gas permeation increased with raising the amount of CDC in the PA layer, the increments were not the same for the two gasses. The maximum permeance of CO2 and CH4 was recorded at 1% loading with a value of 5.00 and 0.375 GPU; respectively. The higher increase in CO2 permeation compared to CH4 was responsible for the improvement in the CO2/CH4 selectivity. Table 6 revealed that the CO2/CH4 selectivity improved by increasing CDC loading up to a concentration of 0.5%, as a result of the excellent dispersion of CDCs in the PA layer. At higher loading the selectivity declined to 13.31. This deterioration in selectivity is attributed to the agglomeration of the nanoparticles observed in the SEM images in Fig. 5. This might have created defects in the membrane and increased the CH4 permeance that resulted in lower CO2/CH4 selectivity.

Table 5 Adsorption uptake of CO2 and CH4 by different materials used in membrane preparation. Materials

CO2 uptake (cc/g)

CH4 uptake (cc/g)

Conditions

Ref

CDC b-CuBDC-MOF CDC Zeolite CDC

125 1.29 (mmol g1) e e 89.78

55 0.182 (mmol g1) 46 35 39.39

298.15 K 273 K and 1 bar 298.15 K 298.15 K 293.15 K

[53] [77] [54] [54] This work

196

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

Table 6 The effect of CDC loading on the gas separation properties at 300.15 K and 5 bar. Gas permeance (GPU)

Gas selectivity

Membrane

Loading%

CO2

CH4

CO2/CH4

MMM0 MMM1 MMM2 MMM3 MMM4 MMM5

0 0.0005 0.002 0.1 0.5 1

2.16 ± 0.04 2.33 ± 0.15 2.44 ± 0.24 3.13 ± 0.28 4.07 ± 0.43 5.00 ± 0.18

0.162 ± 0.001 0.163 ± 0.001 0.168 ± 0.005 0.185 ± 0.001 0.204 ± 0.015 0.375 ± 0.034

13.33 ± 0.36 14.36 ± 0.94 14.51 ± 1.92 16.86 ± 1.44 19.92 ± 3.63 13.31 ± 1.72

3.3.4. The effect of the number of layers The best loading amount (0.5% CDC) was chosen to fabricate CDCs/polyamide MMMs with multiple selective layers (1, 2, 6, and 10 layers)). Table 7 shows the permeance of CO2 and CH4 as a function of the number of selective layers. It can be observed that both the CO2 and CH4 gas permeance were reduced as the number of the layers increased because of the higher mass transfer resistance to gas transport. With 10 layers, the membrane permeance for CO2 decreased 43% while the CH4 permeation decreased 53% compared to that of a single layer. The reduction in the gas permeance was accompanied with a slight enhancement in CO2/CH4 selectivity, wherein the selectivity improved by 21% for ten selective layers. The higher selectivity of the CO2/CH4 is attributed to the functional groups of the CDC/PA layer and to the presence of larger amounts of CDC which maximize the benefits of the CDC structural properties to offer higher permeation of CO2 in comparison with CH4.

3.3.5. The effect of the operating temperature In order to study the effect of temperature on the separation performance, PSF, PA, as well as the CDC/PA membranes were subjected to temperatures varying from 300.15 to 323.15 K. The results are given in Table 8. The temperature dependence of the permeability can be expressed by the Arrhenius equation as follows [79]:

  E P ¼ P0 exp  RT

(4)

operating temperature results in a more flexible polymer chains and higher free volume for gas molecules to transport. This leads to higher gas diffusion and permeation. At the same time, lower gas selectivity is expected by increasing the temperature as a result of the wider polymer chain motions and the looser structure of the membrane [80]. It can be seen in Table 6 that the permeance of all gases was enhanced by increasing the operating temperature. Nevertheless, the permeation improvement was not the same for all gasses. For instance, by increasing the operating temperature from 300.15 to 323.15 K, the permeance of the PA membrane for CO2 increased by 29.9%, while the permeance for CH4 increased by 85.8%. This is due to the fact that CH4 permeation is based on simple molecule diffusion, while CO2 gas has a strong interaction with the functional groups of the PA membrane which results in higher solubility [58,80]. The gas solubility is reduced by increasing the temperature and the reduction is higher for the more condensable gas (CO2) compared to the CH4 gas. For the PSF membrane, CO2 permeance improved by 17.95% as the temperature increased from 300.15 to 323.15 K, while the CH4 permeance increased by 51.1%. The general trend of the gas selectivity showed a decrease as the operating temperature increased. This can be explained by the higher increase in CH4 permeance compared to that of the CO2. The activation energies of CO2 and CH4 were calculated for the PSF, TFC, and TFN membranes and tabulated in Table 9. For the PA and CDC/ PA membranes, the values of the activation energy followed the

Table 9 Activation energies of CO2 and CH4 for PSF, PA, and MMM4 (kJ/mole).

where, P0 is a pre-exponential coefficient, E is the apparent activation energy for gas permeation, R is the gas constant, and T is the absolute temperature. Generally speaking for polymeric membranes, increasing the

Membrane

ECO2 (kJ/mole)

ECH4 (kJ/mole)

PSF PA MMM4

8.04 ± 0.82 9.53 ± 0.06 6.74 ± 0.59

16.32 ± 0.98 18.77 ± 0.63 15.27 ± 0.71

Table 7 The effect of number of layers on the gas separation properties at 300.15 K and 5 bar. Gas permeance (GPU)

Gas selectivity

Number of layers

CO2

CH4

CO2/CH4

1 2 6 10

4.08 ± 0.43 3.37 ± 0.57 3.33 ± 0.18 2.32 ± 0.25

0.204 ± 0.015 0.167 ± 0.003 0.144 ± 0.017 0.096 ± 0.003

19.92 ± 3.63 20.14 ± 2.94 23.10 ± 1.47 24.08 ± 1.60

Table 8 The effect of operating temperature on the separation performance of PSF, PA, and MMM4 at 5 bar. PSF

PA

MMM4

T(K)

CO2

CH4

CO2/CH4

CO2

CH4

CO2/CH4

CO2

CH4

CO2/CH4

300.15 308.15 323.15

2.41 ± 0.15 2.45 ± 0.11 2.85 ± 0.0

0.56 ± 0.01 0.61 ± 0.01 0.84 ± 0.03

4.33 ± 0.77 4.05 ± 0.74 3.38 ± 0.55

2.16 ± 0.07 2.44 ± 0.20 2.80 ± 0.05

0.16 ± 0.02 0.22 ± 0.01 0.30 ± 0.03

13.33 ± 1.90 10.75 ± 1.66 9.29 ± 0.89

4.07 ± 0.43 4.14 ± 0.67 4.89 ± 0.64

0.20 ± 0.01 0.25 ± 0.06 0.32 ± 0.05

19.92 ± 3.63 16.26 ± 1.56 15.32 ± 0.60

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

197

Table 10 The effect of feed-gas pressure on the gas separation properties for PSF, PA, and MMM4. PSF

PA

MMM4

P (bar)

CO2

CH4

CO2/CH4

CO2

CH4

CO2/CH4

CO2

CH4

CO2/CH4

1 2 3.5 5

1.72 ± 0.34 2.23 ± 0.14 2.82 ± 0.26 2.41 ± 0.15

0.583 ± 0.002 0.641 ± 0.001 0.709 ± 0.006 0.556 ± 0.004

2.96 ± 0.65 3.48 ± 0.42 3.98 ± 0.13 4.33 ± 0.77

3.08 ± 0.08 2.85 ± 0.10 2.54 ± 0.02 2.16 ± 0.07

0.151 ± 0.01 0.153 ± 0.25 0.158 ± 0.15 0.162 ± 0.02

20.29 ± 2.10 18.55 ± 1.95 15.98 ± 2.50 13.33 ± 1.90

4.84 ± 0.51 4.77 ± 0.60 4.36 ± 0.66 4.07 ± 0.43

0.159 ± 0.001 0.171 ± 0.006 0.182 ± 0.001 0.204 ± 0.010

30.45 ± 2.25 27.88 ± 3.45 23.87 ± 2.32 19.92 ± 3.63

permeance order in which E (CH4) > E (CO2). The activation energy is generally affected by the gas/polymer interaction as well as the molecular size of the penetrant. In addition, gasses with lower activation energies move faster through the membrane and vice versa. The low activation energy of CO2 can be attributed to the high gas solubility in the PA [79]. Similar observations were reported for polyimide, polyarylene ether, and cellulose membranes [81,82]. 3.3.6. The effect of feed pressure To investigate the effect of the gas feed pressure on the separation performance, different feed pressures were applied to the membranes at a constant temperature of 300.15 K and the results are illustrated in Table 10. It can be seen that CO2 permeance for both the PA and the CDC/PA membranes decreased as the feed pressure increased, while for the PSF membrane there was a slight increase in CO2 permeance. For the PA and CDC/PA membranes, it is known that the CO2 reacts with the functional groups to produce a complex HCO3 e which can diffuse from one site to another. Therefore, by increasing the CO2 amount within the membrane, amine carriers might reach the saturation state where they cannot combine with other CO2 molecules anymore. Consequently, lower CO2 permeance is expected from this mechanism of CO2 permeation. Similar trends have been reported in the literature [83]. However, the enhanced CO2 permeance in the PSF membrane is attributed to the higher molecular concentration of CO2 at the gas/ membrane interface, which increases the driving force and results in enhanced gas permeation. On the other hand, the PA membranepermeance of CH4 was almost constant when the feed pressure increased from 1 to 5 bars, in fact it increased by only 28.6% (for CDC/PA MMMs). Since CH4 molecules do not react with the PA membrane, the transport of CH4 gas across the membrane is based on a simple solutionediffusion mechanism. According to the dual mode sorption model which is common for the glassy polymer [84], at high feed pressure polymer chains become more compact, which results in a lower fractional free volume of polymer and consequently, the gas diffusivity declines. This phenomena could be another reason for the decrease in CO2 permeance. In Table 10, it can be seen that when the feed pressure increased from 1 to 5 bars, the CO2 permeance decreased by 29.8% and 15.8% for PA and CDC/ PA membrane; respectively. The decline in CO2 permeance for the CDC/PA membrane was lower because the CDC nanoparticles provided more free volume for the gas to move and therefore higher permeation is expected. As can be seen in Table 10, the gas selectivity of both the PA and CDC/PA membranes declined with pressure as a result of the reduction in CO2 permeance while maintaining an almost constant CH4 permeance. In contrast, higher feed pressure resulted in a higher molecular concentration at the PSF membrane/gas interface which caused CO2 and CH4 permeance to increase with a slight improvement in the gas selectivity caused by the higher increase in CO2 permeation compared to that of the CH4. 4. Conclusions A novel carbide-derived-carbon composite membrane was

successfully fabricated by IP of piperazine and isophthaloyl chloride using PSF membrane as a support. SEM images confirmed the good distribution of CDC nanoparticles on the PA selective layer without affecting the crystalline structure of the membrane. In addition, the new membrane demonstrated higher thermal stability upon the addition of CDC to the polymeric matrix which absorbed the heat, and consequently suppressed the rate of membrane decomposition. The separation performance of the bare PA membrane was remarkably improved by embedding the CDC nanoparticles into the PA layer because the nanoparticles disrupted the polymeric matrix to offer faster gas diffusion. Moreover, the high surface area (2589 m2/g) and narrow pore size distribution along with the preferential adsorption of CO2 over CH4 enhanced the CO2/CH4 selectivity. By increasing the CDC loading within the polymer, the separation performance improved, although at 1% loading some agglomeration was observed which resulted in lower gas selectivity. Layer-by-layer membranes were prepared and the separation tests demonstrated a declining trend in CO2 and CH4 permeation with a considerable increase in the selectivity. The lower gas permeance is attributed to the increased mass transfer resistance. The present study is an attempt to introduce the CDC nanomaterials as a filler to delve deeper into the thin file nanocomposite membranes for the application of CO2 separation. The results demonstrated the attractive characteristics and enormous potential of CDC for gas separation.

Acknowledgment The authors would like to acknowledge the support provided by King Abdulaziz City for Science and Technology through NSTIP for funding this work under project No. 14-ADV303-04.

Nomenclature

Acronyms MMMs NG PSF PA GPU IP DMACs THF DI IPC PIP TMC MPD FT-IR XRD TGA SEM

Mixed matrix membranes natural gas polysulfone polyamide gas Permeation Units interfacial polymerization n, n-dimethylacetamide tetrahydrofuran deionized water isophthaloyl chloride piperazine trimesoyl chloride m-Phenylenediamine fourier transformation infrared x-ray Diffraction thermogravimetric Analysis scanning Electron Microscopy

198

Variables T R E P2 A Vd

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

cell temperature (K) universal gas constant (kJ/K. kmol) apparent activation energy for gas permeation (kJ/mol) the upstream pressure (kPa) effective membrane area (cm2) downstream volume (cm3)

[23]

[24]

[25]

Greek letter p gas permeance (GPU) pre-exponential coefficient po aij gas selectivity

[26]

[27]

References [28] [1] ExxonMobil, reportOutlook for Energy reports: A View to 2040. http:// corporate.exxonmobil.com/en/energy/energy-outlook/a-view-to-2040. [2] Shimekit B, Mukhtar H. Natural gas purification technologies major advances for CO2 separation and future directions. In: Al-Megren H, editor. Advances in natural gas technology. Croatia: InTech; 2012. p. 235e70. [3] Adewole JK, Ahmad AL, Ismail S, Leo CP. Current challenges in membrane separation of CO2 from natural gas : a review. Int J Greenh Gas Contr 2013;17: 46e65. https://doi.org/10.1016/j.ijggc.2013.04.012. [4] Wang M, Lawal A, Stephenson P, Sidders J, Ramshaw C, Hill W, et al. Postcombustion CO2 capture with chemical absorption : a state-of-the-art review. Chem Eng Res Des 2010;89:1609e24. https://doi.org/10.1016/ j.cherd.2010.11.005. [5] Oyenekan BA, Rochelle GT. Energy performance of stripper configurations for CO2 capture by aqueous amines. Ind Eng Chem Res 2006;45:2457e64. [6] Zhang S, Shen Y, Shao P, Chen J, Wang L. Kinetics, thermodynamics, and mechanism of a novel biphasic solvent for CO2 capture from flue gas. Environ Sci Technol 2018;52:3660e8. [7] Wang L, An S, Li Q, Yu S, Wu SI. Phase change behavior and kinetics of CO2 absorption into DMBA/DEEA solution in a wetted-wall column. Chem Eng J 2017;314:681e7. [8] Li Q, Wang YI, An S, Wang L. Kinetics of CO2 absorption in concentrated K2CO3/ PZ mixture using a wetted-wall column. Energy Fuels 2016;30. 7496e02. [9] Ho MT, Allinson GW, Wiley DE. Reducing the cost of CO2 capture from flue gases using membrane technology. Ind Eng Chem Res 2008;47:1562e8. [10] Tuinier MJ, Annaland MVS, Kramer GJ, Kuipers JAM. Cryogenic CO2 capture using dynamically operated packed beds. Chem Eng Sci 2010;65:114e9. [11] Zhang Y, Sunarso J, Liu S, Wang R. Current status and development of membranes for CO2/CH4 separation : a review. Int J Greenh Gas Contr 2013;12: 84e107. [12] Aroon MA, Ismail AF. Performance studies of mixed matrix membranes for gas separation : a review. Separ Purif Technol 2010;75:229e42. [13] Rufford TE, Smart S, Watson GCY, Graham BF, Boxall J, Diniz JC, et al. The removal of CO 2 and N 2 from natural gas : a review of conventional and emerging process technologies. J Petrol Sci Eng 2012;94e95:123e54. https:// doi.org/10.1016/j.petrol.2012.06.016. [14] Rezakazemi M, Ebadi A. State-of-the-art membrane based CO2 separation using mixed matrix membranes ( MMMs ): an overview on current status and future directions. Prog Polym Sci 2014;39:817e61. https://doi.org/10.1016/ j.progpolymsci.2014.01.003. [15] Favre E. Membrane processes and postcombustion carbon dioxide capture : challenges and prospects. Chem Eng J 2011;171:782e93. https://doi.org/ 10.1016/j.cej.2011.01.010. [16] Kim S, Chen L, Johnson JK, Marand E. Polysulfone and functionalized carbon nanotube mixed matrix membranes for gas separation : Theory and experiment. J Membr Sci 2007;294:147e58. https://doi.org/10.1016/ j.memsci.2007.02.028. [17] Chung T, Ying L, Li Y, Kulprathipanja S. Mixed matrix membranes (MMMs) comprising organic polymers with dispersed inorganic fillers for gas separation. Prog Polym Sci 2007;32:483e507. https://doi.org/10.1016/ j.progpolymsci.2007.01.008. [18] Dong G, Chen V. Challenges and opportunities for mixed-matrix membranes for gas separation. J Mater Chem 2013;1:4610e30. https://doi.org/10.1039/ c3ta00927k. [19] Belhaj S, Takagaki A, Sugawara T, Kikuchi R, Oyama ST. Mixed matrix membranes using SAPO-34/polyetherimide for carbon dioxide/methane separation. Separ Purif Technol 2015;148:38e48. https://doi.org/10.1016/ j.seppur.2015.04.017. [20] Bastani D, Esmaeili N, Asadollahi M. Polymeric mixed matrix membranes containing zeolites as a filler for gas separation applications : a review. J Ind Eng Chem 2013;19:375e93. https://doi.org/10.1016/j.jiec.2012.09.019. €ktu M. Sod -ZMOF/Matri[21] Atalay-oral Ç, Sirkecio A, Tantekin-ersolmaz S¸B, Go mid s mixed matrix membranes for CO2 separation. J Membr Sci 2015;489: 81e9. [22] Loloei M, Omidkhah M, Moghadassi A. Preparation and characterization of Matrimid ® 5218 based binary and ternary mixed matrix membranes for CO2

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

separation. Int J Greenh Gas Contr 2015;39:225e35. https://doi.org/10.1016/ j.ijggc.2015.04.016. Amooghin AE, Kargari A. Enhanced CO2 transport properties of membranes by embedding nano-porous zeolite particles into. RSC Adv 2015;5:8552e65. https://doi.org/10.1039/C4RA14903C. Dorosti F, Omidkhah M, Abedini R. Enhanced CO2/CH4 separation properties of asymmetric mixed matrix membrane by incorporating nano-porous ZSM-5 and MIL-53 particles into Matrimid ® 5218. J Nat Gas Sci Eng 2015;25:88e102. https://doi.org/10.1016/j.jngse.2015.04.033. Arjmandi M, Pakizeh M. Mixed matrix membranes incorporated with cubicMOF-5 for improved polyetherimide gas separation membranes : Theory and experiment. J Ind Eng Chem 2014;20:3857e68. https://doi.org/10.1016/ j.jiec.2013.12.091. Li J, Yang J. Separation of CO2/CH4 and CH4/N2 mixtures using MOF-5 and Cu3 ( BTC ) 2. J Energy Chem 2014;23:453e60. https://doi.org/10.1016/S20954956(14)60171-6. Huang L, Wang H, Chen J, Wang Z, Sun J, Zhao D, et al. Synthesis, morphology control, and properties of porous metal e organic coordination polymers. Microporous Mesoporous Mater 2003;58:105e14. Zornoza B, Tellez C, Coronas J, Gascon J, Kapteijn F. Microporous and Mesoporous Materials Metal organic framework based mixed matrix membranes ;: an increasingly important field of research with a large application potential. Microporous Mesoporous Mater 2013;166:67e78. https://doi.org/10.1016/ j.micromeso.2012.03.012. Zulhairun AK, Ismail AF. The role of layered silicate loadings and their dispersion states on the gas separation performance of mixed matrix membrane. J Membr Sci 2014;468:20e30. https://doi.org/10.1016/ j.memsci.2014.05.038. Wu H, Li X, Li Y, Wang S, Guo R, Jiang Z, et al. Facilitated transport mixed matrix membranes incorporated with amine functionalized MCM-41 for enhanced gas separation properties. J Membr Sci 2014;465:78e90. https:// doi.org/10.1016/j.memsci.2014.04.023. Park S, Bang J, Choi J, Hyup S, Lee J, Suk J. 3-Dimensionally disordered mesoporous silica ( DMS ) -containing mixed matrix membranes for CO2 and nonCO2 greenhouse gas separations. Separ Purif Technol 2014;136:286e95. https://doi.org/10.1016/j.seppur.2014.09.016. Xin Q, Gao Y, Wu X, Li C, Liu T, Shi Y, et al. Incorporating one-dimensional aminated titania nanotubes into sulfonated poly ( ether ether ketone ) membrane to construct CO2 -facilitated transport pathways for enhanced CO2 separation. J Membr Sci 2015;488:13e29. https://doi.org/10.1016/ j.memsci.2015.02.047. Xin Q, Wu H, Jiang Z, Li Y, Wang S. SPEEK/amine-functionalized TiO2 submicrospheres mixed matrix membranes for CO2 separation. J Membr Sci 2014;467:23e35. https://doi.org/10.1016/j.memsci.2014.04.048. Zhou T, Luo L, Hu S, Wang S, Zhang R, Wu H. Janus composite nanoparticleincorporated mixed matrix membranes for CO2 separation. J Membr Sci 2015;489:1e10. https://doi.org/10.1016/j.memsci.2015.03.070. Li X, Jiang Z, Wu Y, Zhang H, Cheng Y. High-performance composite membranes incorporated with carboxylic acid nanogels for CO2 separation. J Membr Sci 2015;495:72e80. https://doi.org/10.1016/j.memsci.2015.07.065. Li X, Wang M, Wang S, Li Y, Jiang Z, Guo R, et al. Constructing CO2 transport passageways in Matrimid s membranes using nanohydrogels for efficient carbon capture. J Membr Sci 2015;474:156e66. https://doi.org/10.1016/ j.memsci.2014.10.003. Saeid S, Li Y, Chung T, Liu Y. Enhanced gas separation performance of nanocomposite membranes using MgO nanoparticles. J Membr Sci 2007;302: 207e17. Murali RS, Padaki M, Matsuura T, Abdullah MS, Ismail AF. Polyaniline in situ modified halloysite nanotubes incorporated asymmetric mixed matrix membrane for gas separation. Separ Purif Technol 2014;132:187e94. https:// doi.org/10.1016/j.seppur.2014.05.020. Li X, Ma L, Zhang H, Wang S, Jiang Z, Guo R, et al. Synergistic effect of combining carbon nanotubes and graphene oxide in mixed matrix membranes for efficient CO2 separation. J Membr Sci 2015;479:1e10. https:// doi.org/10.1016/j.memsci. 2015.01.014. Ansaloni L, Zhao Y, Jung BT, Ramasubramanian K, Giacinti M, Ho WSW. Facilitated transport membranes containing amino-functionalized multiwalled carbon nanotubes for high-pressure CO2 separations. J Membr Sci 2015;490:18e28. Ranjbaran F, Omidkhah MR, Amooghin AEJ. The novel Elvaloy4170/functionalized multi-walled carbon nanotubes mixed matrix membranes : fabrication, characterization and gas separation study. J Taiwan Inst Chem Eng 2015;49:220e8. https://doi.org/10.1016/j.jtice.2014.11.032. Nasir R, Mukhtar H, Man Z, Shima M. Development and performance prediction of polyethersulfone-carbon molecular sieve mixed matrix membrane for CO2/CH4 separation. J Memb Sci 2015;45:1417e22. https://doi.org/ 10.3303/CET1545237. Wong KC, Goh PS, Ismail AF. Thin film nanocomposite : the next generation selective membrane for CO2 removal. J Mater Chem A Mater Energy Sustain 2016;4:15726e48. https://doi.org/10.1039/C6TA05145F. Albo J, Wang J, Tsuru T. Gas transport properties of interfacially polymerized polyamide composite membranes under different pre-treatments and temperatures. J Membr Sci 2014;449:109e18. https://doi.org/10.1016/ j.memsci.2013.08.026. Zhang Y, Xiao C, Liu E, Du Q, Wang X. Investigations on the structures and

A. Awad, I.H. Aljundi / Energy 157 (2018) 188e199

[46]

[47]

[48]

[49]

[50] [51]

[52]

[53]

[54]

[55] [56]

[57]

[58]

[59]

[60] [61]

[62] [63]

[64] [65]

[66]

performances of a polypiperazine amide/polysulfone composite membrane. J Membr Sci 2006;191:291e5. https://doi.org/10.1016/j.desal.2005.08.016. Shepl YA, Kotp YH, El-deab MS, Shawky HA, El-anadouli BE. Performance enhancement of TFC RO membrane by using magnesium silicate nanoparticles. Desalination 2012;191:291e5. Emadzadeh D, Lau WJ, Rahbari-sisakht M, Daneshfar A, Ghanbari M, Mayahi A. A novel thin film nanocomposite reverse osmosis membrane with superior anti-organic fouling affinity for water desalination. Desalination 2015;368: 106e13. Lind ML, Suk DE, Hoek EMV. Tailoring the structure of thin film nanocomposite membranes to achieve seawater RO membrane performance. Environ Sci Technol 2010;44:8230e5. Lianchao L, Baoguo W, Huimin T, Tianlu C, Jiping X. A novel nanofiltration membrane prepared with PAMAM and TMC by in situ interfacial polymerization on PEK-C ultrafiltration membrane. J Membr Sci 2006;269:84e93. Kim J, Lee K, Youl S. Pervaporation separation of water from ethanol through polyimide composite membranes. J Membr Sci 2000;169:81e93. Sevilla M, Mokaya R. Activation of carbide-derived carbons : a route to materials with enhanced gas and energy storage properties. J Mater Chem 2011: 4727e32. Yushin GN, Hoffman EN, Nikitin A, Ye H, Barsoum MW, Gogotsi Y. Synthesis of nanoporous carbide-derived carbon by chlorination of titanium silicon carbide. Carbon 2005;43:2075e82. https://doi.org/10.1016/j.carbon.2005.03.014. Gogotsi Y, Fischer JE. Nanoporous carbide derived carbon with tunable pore size nanoporous carbide derived carbon with tunable pore size. Nat Mater 2003;2:591e4. Dash R, Chmiola J, Laudisio G. Titanium carbide derived nanoporous carbon for energy-related applications titanium carbide derived nanoporous carbon for energy-related applications. Carbon 2006;44:2489e97. Hoffman EN, Yushin G, Wendler BG, Barsoum MW, Gogotsi Y. Carbide-derived carbon membrane. Mater Chem Phys 2008;112:587e91. Hoffman EN. Carbide derived carbon from max-phases and their separation applications. Pennsylvania: A Thesis Submitted to the Faculty of Drexel University; 2006. Khan A, Khan Z, Aljundi IH. Improved hydrophilicity and anti-fouling properties of polyamide TFN membrane comprising carbide derived carbon. Desalination 2017;420:125e35. https://doi.org/10.1016/j.desal.2017.07.002. Rabiee H, Meshkat S, Soltanieh M. Gas permeation and sorption properties of mixed matrix membrane for CO2/CH4 and CO2/N2 separation. J Ind Eng Chem 2015;27:223e39. https://doi.org/10.1016/j.jiec.2014.12.039. Jan P, Detlev F, Thomas K, Klaus-viktor P. Gas permeation measurement under defined humidity via constant volume/variable pressure method. J Membr Sci 2012;389:343e8. https://doi.org/10.1016/j.memsci.2011.10.046. Lin W, Vora RH, Chung T. Gas transport properties of 6FDA-durene/1,4phenylenediamine ( pPDA ) copolyimides. J Membr Sci 2000;389:2703e13. Chi M, Cong H, Yu B, Yuan H, Luo Y, et al. Carbon nanotube/brominated poly ( 2, 6- dimethyl-1, 4-phenylene oxide ) nanocomposite membranes for CO2/N2 separation. Integr Ferroele 2015;164:136e44. Gogotsi Y. Carbon nanomaterials handbook. Pennsylvania: Taylor and Francis group, LLC. CRC Press; 2006. Sing K, Everett D, Haul R, et al. Reporting physisorption data for gas/solid systems with special reference to the determination of surface area and porosity. Pure Appl Chem 1985;57:603e19. Socrates G. Infrared and Raman characteristic group frequencies. third ed. West Sussex: John Wiley & Sons, Lts; 2001. Zera E, Nickel W, Hao GP, Vanzetti L, Kaskel S, Sorar GD. Nitrogen doped carbide derived carbon aerogels by chlorine etching of a SiCN aerogel. J Mater Chem A 2016:4525e33. https://doi.org/10.1039/C6TA00589F. Adu KW, Li Q, Desai SC, Sidorov AN, Sumanasekera GU, Lueking AD.

[67] [68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78] [79]

[80]

[81]

[82] [83]

[84]

199

Morphological, structural, and chemical effects in response of novel carbide derived carbon sensor to NH3, N2O, and air. Langmuir 2009;25:582e8. Pesek SC, Koros WJ. Aqueous quenched asymmetric polysulfone membranes prepared by dry/wet phase separation. J Membr Sci 1993;81:71e88. Aroon MA, Ismail AF, Montazer-rahmati MM, Matsuura T. Morphology and permeation properties of polysulfone membranes for gas separation : effects of non-solvent additives and co-solvent. Separ Purif Technol 2010;72: 194e202. https://doi.org/10.1016/j.seppur.2010.02.009. Chakrabarty B, Ghoshal AK, Purkait MK. Preparation, characterization and performance studies of polysulfone membranes using PVP as an additive. J Membr Sci 2008;315:36e47. https://doi.org/10.1016/j.memsci.2008.02.027. Boom RM, Wlenk IM, Boomgaard T Van Den, Smolders CA. Microstructures in phase inversion membranes. Part 2. The role of a polymeric additive. J Membr Sci 1992;73:277e92. Online VA, Wang M, Wang Z, Li S, Zhang C, Wang J, et al. A high performance antioxidative and acid resistant membrane prepared by interfacial polymerization for CO2 separation from flue gas. Energy Environ Sci 2013;6:539e51. https://doi.org/10.1039/c2ee23080a. Bruck SD. Thermal degradathgn of piperazine polyamides H poly ( terephthaloyl piperazine ), poly ( terephthaloyl 2-methylpiperazine ) and poly ( isophthaloyl trans-2,5-dimethylpiperazine). Polymer 1966;7:231e41. Liang C, Uchytil P, Petrychkovych R, Lai Y, Friess K, Sipek M, et al. A comparison on gas separation between PES ( polyethersulfone )/MMT ( Namontmorillonite ) and PES/TiO2 mixed matrix membranes. Separ Purif Technol 2012;92:57e63. Sridhar S, Smitha B, Mayor S, Prathab B, Aminabhavi T, et al. Gas permeation properties of polyamide membrane prepared by interfacial polymerization. J Mater Sci 2007;42:9392e401. https://doi.org/10.1007/s10853-007-1813-5. Wong KC, Goh PS, Ng BC, Ismail AF. Thin film nanocomposite embedded with polymethyl methacrylate modified multi-walled carbon nanotubes for CO2 removal. RSC Adv 2015;5:31683e90. https://doi.org/10.1039/C5RA00039D. Yu X, Wang Z, Wei Z, Yuan S, Zhao J. Novel tertiary amino containing thin film composite membranes prepared by interfacial polymerization for CO2 capture. J Membr Sci 2010;362:265e78. https://doi.org/10.1016/ j.memsci.2010.06.043. Erucar I, Keskin S. High CO2 selectivity of an amine-functionalized metal organic framework in adsorption-based and membrane-based gas separations. Ind Eng Chem Res 2013;52:3462e72. Li J, Kuppler R, Zhou H. Selective gas adsorption and separation in metaleorganic frameworks. Chem Soc Rev 2009;38:1477e504. Sen SK, Banerjee S. Gas transport properties of fluorinated poly ( ether imide ) films containing phthalimidine moiety in the main chain. J Membr Sci 2010;350:53e61. https://doi.org/10.1016/j.memsci.2009.12.011. Bera D, Bandyopadhyay P, Ghosh S, Banerjee S. Gas transport properties of aromatic polyamides containing adamantyl moiety. J Membr Sci 2014;453: 175e91. Banerjee S, Maier G, Dannenberg C, Spinger J. Gas permeabilities of novel poly ( arylene ether)s with terphenyl unit in the main chain. J Membr Sci 2004;229:63e71. Li X, Kresse I, Xu Z. Effect of temperature and pressure on gas transport in ethyl cellulose membrane. Polymer 2001;42:6801e10. Wang M, Wang Z, Li N, Liao J, Zhao S, Wang J, et al. Relationship between polymer e filler interfaces in separation layers and gas transport properties of mixed matrix composite membranes. J Membr Sci 2015;495:252e68. https:// doi.org/10.1016/j.memsci.2015.08.019. Ebadi A, Omidkhah M, Kargari A. The effects of aminosilane grafting on NaY zeolite e matrimid s 5218 mixed matrix membranes for CO2/CH4 separation. J Membr Sci 2015;490:364e79. https://doi.org/10.1016/ j.memsci.2015.04.070.