Fabrication of nanohybrid polyetherimide/graphene oxide membranes: biofuel dehydration by pervaporation process

Abstract

Polyetherimide (PEI) flat-sheet membranes (PM) as a nanohybrid membrane were fabricated through dry-thermal (PMDT) and thermal treatment (PMTT) methods using graphene oxide (GO) as a nanofiller. Characterization of PM membranes was carried out using Fourier transform infrared (FTIR), scanning electron microscopy (SEM) and X-ray diffraction (XRD). The PMDT membrane cross-section showed a sponge-like structure resulting in higher permeability and lower separation factor. The SEM image (cross-section) of the PMTT membrane showed a dense structure with higher separation factor and lower permeation flux than those of the PMDT membrane. Incorporation of GO into the PEI matrix significantly reduced the contact angle surface hydrophobicity of the membranes. The nanohybrid membrane exhibited an exceptional pervaporation performance at a concentration of 95 wt% n-butanol. The permeation flux of PMDT and PMTT membranes were 1100.26 and 642.69 kg m⁻² h⁻¹ and their separation factors were 89.39 and 96.96, respectively. The GO incorporation and novel fabrication techniques resulted in excellent separation performance.

---

Introduction

Pervaporation (PV) is a membrane process that is applied solely for liquid phase separation, especially for azeotropic and close boiling mixtures. The advantages of separation processes through the pervaporation membranes in comparison with the prevalent separation processes, such as distillation, reverse osmosis and ultrafiltration, are low energy consumption and minimum operating costs.^1–3 In pervaporation, the membrane surface is in direct contact with the feed solution. The solution-diffusion model describes separation in the PV process in three steps of dissolution onto the membrane, diffusion through the membrane and evaporation to the down-stream side which called permeation.

Basically, hydrophilic and hydrophobic polymeric membranes are used to separate water from the organic solvents.^4–7 Typically, hydrophilic membranes are considered to dehydrate n-butanol as a biofuel, which is considered as a future energy source to substitute with gasoline.^8–14 The heating value of n-butanol (2501 kJ mol⁻¹) is higher than that of ethanol (1329 kJ mol⁻¹); for this reason, this alcohol is used as a biofuel and blended with gasoline and diesel. The combustion of this biofuel resulted in the production of water and carbon dioxide as well.^15,16 Since fossil fuels have natural organic sources, the presence of gases with sulfur and nitrogen atoms is expected. Therefore, in addition of CO₂ and H₂O formation, the gases of SO_x, NO_x and CO are produced during the fuel combustion process owing to fuel impurities. Production of CO is related to a defect in fuel combustion. NO_x emissions were reduced because of its higher heat of evaporation, which resulted in a lower ignition temperature. Combustion of n-butanol increases the temperature to assist the fuel combustion. The results showed that adding n-butanol has more advantages than the widely used ethanol. n-Butanol has a 53% greater heating value than that of ethanol.^17,18 n-Butanol (C₄H₉OH) is an alcohol with a four-carbon chain as a non-polar group and a polar hydroxyl group, which is responsible for its solubility in water and polar solvents. The solubility of this alcohol in water (73 g L⁻¹) is less than the lighter alcohols with lower carbon atoms, such as propanol, ethanol and methanol, owing to its larger non-polar group. These light alcohols are fully miscible in water. Therefore, n-butanol shows limited solubility in comparison with the light alcohols. n-Butanol is produced from the petrochemical industry and fermentation, where it is used as an industrial intermediate and a biofuel for diesel fuel and gasoline. The n-butanol that is obtained from the fermentation process has less than 3.0 wt% concentration of ABE products (acetone, butanol and ethanol).¹⁹ The separation process of n-butanol involves two main steps: purification of the dilute solution by distillation to produce a solution of 80 wt% n-butanol and dehydration of the purified solution to concentrate it further to 99.5 wt%. Although the recovery of n-butanol by using membranes shows good separation performance, it is not economically feasible for large-scale industrial applications. Mostly polymeric membranes are not able to separate the water/organic solutions with good results owing to the swelling and cracking of the membranes and the low permeation process. Moreover, though many membranes have been used to dehydrate biofuels with high performance, they are not economical or easily fabricated to use in large-scale in different industries. Some materials, methods and techniques, such as fillers, grafting, cross-linking, support-layer^20,21 and coating,^22–24 have been used to overcome the current issues with the polymeric materials. Polyetherimide (PEI) with specific properties such as excellent mechanical/chemical properties, less swelling characteristics, solvent resistance, good membrane-forming property, and high separation performance²⁵ is a suitable candidate as a matrix polymer for dehydration of n-butanol. Therefore, this polymer with the attractive features was used for separation of n-butanol from its aqueous solution. Since PEI does not have a significant permeation performance, it should be modified by using graphene oxide (GO) to enhance the permeation process by increasing the hydrophilicity of the membrane.^26–33

Recently, graphene has attracted researchers because of its derivative graphene oxide (GO), which exhibits different characteristics in comparison with graphene. Graphene has been used to prepare a film with highly hydrophobic property containing multilayer graphene platelets, while graphene oxide (GO) is a hydrophilic material with a wide range of active oxygen functional groups, such as hydroxyl, carboxyl and epoxide groups, which are responsible for its polar properties. Therefore, it can be rendered as a good compatible material with PEI to prepare hydrophilic membranes.^26–29 These functional groups improved the interfacial interaction between graphene oxide and the polymer matrix, which paved the way for the development of the GO based nano-composites with significant mechanical and thermal properties. Nano-GO is a hydrophilic material with numerous active oxygenated groups that can be used for the dehydration process. The polar groups of nano-GO can easily interact with water molecules and lead to a fast diffusion rate in the permeate side and high selectivity towards water molecules.²⁹ Here, the hydrophilicity of the membranes is increased and promotes the favorable properties of the membranes for the dehydration of organic solvents. Generally, modification with inorganic materials shows low performance in the separation factor, while nano-graphene oxide offers higher performance in the separation process.^9,34–39

The application of GO is based on the number of hydroxyl groups of the GO network, which indicates the presence of the large specific surface area, superior mechanical stiffness and flexibility, remarkable optical transmittance, exceptionally high electronic and thermal conductivities, impermeability to gases and water, as well as many other supreme properties. These properties make it a miracle material. Scheme 1 shows the chemical structure of GO.

Scheme 1 The chemical structure of graphene oxide.

Most of the conventional membranes are not economical and are not easy to fabricate due to the expensive modifications and preparation processes required.^40–43 They are not able to achieve the target of simultaneously improving permeation and separation factor. This study involved the fabrication of hydrophilic PMDT (dry-thermal) and PMTT (thermal treatment) membranes by using two different casting techniques to achieve different micro-structure membranes with different pervaporation performances. The PMDT and PMTT membranes possess significant features, such as an easy and low cost fabrication process without using any modifications, such as grafting and coating. There were no modification techniques in the fabrication processes. The PMDT and PMTT membranes were prepared easily by incorporating nano-graphene oxide to enhance the favorable properties of the permeation process. In addition, the different morphological structures, separation performances and effects of nano-fillers such as nano-graphene oxide in the activity of these membranes were investigated in detail. The nano-filler structure was studied by using XRD patterns and TEM images.

Method and materials

Nano-graphene oxide as a filler in the membrane was synthesized from graphene powder with an average particle thickness of approximately 6 to 8 nanometers and a typical surface area of 120 to 150 m² g⁻¹. Graphene powder was purchased from XG Sciences, Inc. Polyetherimide (PEI) (C₃₇H₂₄O₆N₂)_n utilized as the base material of the membrane with a density of 1.27 g cm⁻³ was provided by Ultem, General Electric Company. Concentrated sulfuric acid (H₂SO₄, 98 wt%), potassium permanganate (KMnO₄), sodium nitrate (NaNO₃) and hydrochloric acid (HCl) were used (Merck) and hydrogen peroxide (H₂O₂, 30 wt%) and N-methyl-2-pyrrolidone (NMP) were purchased from QREK and J. T. Baker, respectively.

Preparation of graphene oxide

Typically, the preparation of graphene oxide was carried out in three stages. In the first stage, graphene powder (2 g), sodium nitrate (1 g) and 46 mL of concentrated sulfuric acid (H₂SO₄ 98%) were added into a reactor and stirred to control the temperature. In the second stage, KMnO₄ (8 g) was added gradually to the mixture at temperature below 10 °C for 2 h, and then the temperature was increased to 35 °C and stirred for 1 h. In the third stage, the mixture was diluted with deionized water (96 mL) with the temperature below 100 °C and stirred for 1 h followed by further diluting with deionized water (100 mL). After that, 4 mL of 30% H₂O₂ was added to the mixture to reduce the residual KMnO₄ and the mixture color changed into a brilliant yellow. Finally, the mixture was centrifuged and washed with 5% HCl aqueous solution. The wet powder was dried using a freeze dryer and a fine coffee color powder was obtained at the end of this process. The chemical reactions of these three steps are shown below (eqn (1)–(3)):

NaNO₃ + H₂SO₄ → HNO₃ + NaHSO₄ (1)

2KMnO₄ + H₂SO₄ → K₂SO₄ + Mn₂O₇ + H₂O (2)

Dihydroxylation in presence of water (3)

Membrane preparation

Hybrid membranes (PMDT and PMTT) were prepared by mixing PEI (15 wt%) and GO (0.5 wt%) and NMP (84 wt%) as a common solvent. Sonication was used to avoid agglomeration and degassing of the solvent before casting. The first step required a uniform dispersion of GO in NMP by using sonication for 60 minute at 50–60 °C. Then the solvent was stirred while polyetherimide was added to the mixture by a priming technique at 60 to 70 °C at 600 rpm for 12 h to dissolve the polymer completely in the solvent. The mixture required sonication to remove bubbles from the mixture. Before casting, the solvent was kept at room temperature for 1 h to reach equilibrium temperature with the lab environment to avoid a fast phase transition between the solvent and room temperature.

Membrane casting

Two techniques were employed for casting the flat sheet hybrid membranes: dry-thermal (PMDT) and thermal-treatment (PMTT). The PMDT membrane was prepared into two stages; the first stage is casting the solvent on the flat glass. The flat glass membrane sheet is kept at room temperature for 1 h, which allows slow evaporation of the surface layer of the membrane. Then this flat sheet was dried in the oven for 12 h at 40 °C. The PMTT membrane was kept in the oven immediately after casting on glass sheet for 1 h at 50 °C and then the temperature was increased to 60 °C for 6 h, after which the PMTT membrane was ready for the next stage.

Pervaporation experiments

The separation performance of the dense blended GO/PEI membranes was determined through pervaporation experiments. Feed solution was pumped across the membrane surface with an effective area of 78 cm². A vacuum pump was set at 0.05 bar to keep the pressure of the membrane permeated side and the permeated vapor was collected in an ice bath/cold trap. The permeation flux (J, kg m⁻² h⁻¹) was obtained according to the eqn (4), the separation factor of the permeated product was analyzed by the refractometer index and the result of the analysis was calculated from eqn (5):

J = Q/At (4)

α = (Y_water/Y_butanol)/(X_water/X_butanol) (5)

where Q (g) is the mass of the permeated product during the operation time t (h), the effective membrane area is A (m²) and X and Y are the weight fractions of species in the feed and permeate, respectively.^45,47

Characterization of hydrophilic membrane

XRD analysis was used to determine the oxidization degree of GO. Transmission electron microscopy (TEM, Tecnai G2T-20S, 2004) was utilized to show the exfoliation level of GO. The surface and cross-section morphologies of the membranes were observed with a scanning electron microscope (SEM, HITACHI, TM 3000). Fourier transform infrared spectroscopy (FTIR, NICOLET-iS5) was used to identify the presence of GO and the interaction between the GO and polymer. The surface hydrophilic behavior of the membrane surface was determined by water contact angle measurement. To minimize experimental error, an average value of the contact angle was obtained by using 15 points of samples. The long-term stability of PMDT and PMTT membranes was investigated for 12 h at a constant temperature of 50 °C.

Results and discussion

Characterization of GO and PEI/GO membranes

Fig. 1A shows the XRD patterns of graphene oxide. The results show an intense diffraction peak at 2θ = 10.5 degrees, which indicates the presence of a typical two-dimensional monolayer structure and the oxidation degree of graphene. In addition, the peak corresponds to an interlayer spacing (d-spacing) hybridization. GO reveals a wide d-spacing due to the numerous oxygen-containing groups, such as hydroxyl, carboxyl, and epoxide, of the former. These functional groups were formed as a result of the oxidation process on the latter, which formed the distance between the GO layers.^46–48

Fig. 1 XRD pattern and TEM images of graphene oxide.

In addition, the well synthesized monolayer graphene oxide was indicated by using transmission electron microscopy (TEM), which is shown in Fig. 1. The surface morphology and cross section of the PMTT and PMDT membranes were evaluated by the representative SEM image, which are shown in Fig. 2.

Fig. 2 The surface morphology SEM of PMTT (A), PMDT (C), cross section SEM of PMTT (B and B-1) and PMDT (D and D-1) membranes (insets are high magnification images and red arrows denote GO sheets).

Fig. 2A and C show the surface morphology and SEM cross-section B, B-1 and D, D-1 of PMTT and PMDT, respectively. The SEM images show a dense structure for PMTT with a higher separation factor and a lower permeation performance, and exhibited a lower density structure for PMDT with a higher permeation performance for the PV process. These images exhibited the effects of the temperature and drying time on their morphologies. Both of these factors affected the physical (density degree) and chemical (hydrophilicity) properties of the membranes. The SEM results exhibited a physical difference between PMDT with low density and PMTT with high density microstructures, which resulted in the difference in their PV performances. PMDT exhibited a higher permeation due to its lower microstructure density, while PMTT exhibited a lower permeation due to its high microstructure density and its separation factor performance was higher than that of the PMDT membrane as well. The low density PMDT membrane was synthesized by using low temperature and slow solvent evaporation (drying time), which resulted in an improvement in the membrane permeation performance. The high density PMTT membrane was prepared by using high temperature and fast solvent evaporation (thermal treatment), which resulted in a lower permeation and a higher separation factor.

The uniform and smooth surface areas of the PMDT and PMTT membranes are shown in Fig. 3, which indicated the uniform dispersion of GO as a filler in the PEI polymer. The agglomeration of GO was not observed on the membrane sheet, which confirmed the uniform dispersion of GO into the PEI polymer.

Fig. 3 Photographs of the PMDT and PMTT membranes.

The observation of any agglomerated GO indicated the leaching or cracking on the surface during the separation process .⁴⁹

Fig. 4 shows the FTIR spectra of the GO and the PMDT and PMTT composite membranes. To verify bonding of the PEI and GO as a nanohybrid membrane, this spectroscopy was used. The characteristic absorption bands of GO and these polymer membranes can be observed at around 3050–3450 cm⁻¹ (O–H bonds of graphene oxide), 2960 and 2850 cm⁻¹ (C–H str), 1716 and 1650 cm⁻¹ (C=O str), 1364 cm⁻¹ (C–N str) and 1050 cm⁻¹ (C–O str).⁴⁴ The results strongly confirmed the synthesis of two types of membranes by observing of two bonds at 1716, and 1364 cm⁻¹, which are evidence of the presence of imide groups (O=C–N–C=O) in their structures.

Fig. 4 FTIR spectra of GO, PMDT and PMTT composite membranes.

The hydrogen bonds due to the interaction between OH groups on the surface of GO with polar groups on the composites shifted the hydrogen broad band from 3290 cm⁻¹ in graphene oxide to 3350 cm⁻¹ in the membranes. This indicates that the surface OH groups of GO have been plugged by the PEI polymer, which results in a decrease in the hydrogen bonding due to the deposition of polymer.

Fig. 5 shows the permeation fluxes and separation factors of PMDT and PMTT at different temperatures of 30, 40 and 50 °C. The results show that the physical and chemical properties of these membranes for dehydration of n-butanol solution are different due to two main factors, namely temperature and drying time. The low density PMDT was synthesized by using low temperature and slow solvent evaporation (drying time), which resulted in an improvement in the membrane permeation performance. The high density PMTT membrane was prepared by using high temperature and fast solvent evaporation (thermal treatment), which resulted in low permeation and a high separation factor.

Fig. 5 Permeation flux (A) and separation factor (B) of PMDT and PMTT membranes at 30, 40 and 50 °C.

Pervaporation performance of PMDT and PMTT composite membranes

Pervaporation experiments were carried out with constant concentrations of 95% n-butanol and 5% water at different temperatures of 30, 40, 50 °C by the two novel nanohybrid membranes: dry-thermal membrane (PMDT) and thermal-treatment membrane (PMTT). The results indicated that PMDT and PMTT membranes displayed exceptional pervaporation performances in this solution with different temperatures of 30, 40, 50 °C. The PMDT showed higher permeation than that of PMTT. The permeation flux in PMDT was reduced by temperature from 1100.26 at 30 °C to 669 kg m⁻² h⁻¹ at 50 °C at a constant pressure of 0.05 bar. Meanwhile, the highest separation factor of PMDT was 89.39 at 40 °C.

On the contrary, the permeation flux of the PMTT membrane was increased with temperature from 419.36 to 642.69 kg m⁻² h⁻¹ at 30 °C and 50 °C, respectively (Fig. 5A). However, the PMTT membrane exhibited a constant separation factor at 40 and 50 °C (Fig. 5B).

The results showed that the PMDT membrane had higher permeation and a lower separation factor owing to the morphological structure, which showed an asymmetric lower density structure, which led to higher diffusion of water molecules. Polyetherimide is relatively hydrophobic with a high contact angle. Table 1 shows that the contact angles of the PMDT and PMTT membranes decreased with adding GO and fabrication techniques, which resulted in an increase in the hydrophilicity property. The presence of graphene oxide, which possesses functional groups, improved the hydrophilicity of the membrane, which resulted in the formation of a greater flux to water molecules.

Table 1 Contact angle of PMDT and PMTT membranes

Membrane	Neat PEI (PMTT)	Neat PEI (PMDT)	Thermal-treatment	Dry-thermal
Contact angle (degree)	77	58	66	43

---

The GO incorporation into the PEI matrix decreased the contact angle of the membrane. Incorporating nano-GO in the membrane increased the hydrophilicity of the membranes. Hence, hybrid membranes were appropriate for dehydration application. The novel modified techniques were applied for casting of the low (PMDT) and high density (PMTT) membranes for dehydration of n-butanol mixtures.⁴²

Fig. 6 shows the long-term performance of the PMDT and PMTT membranes over 120 h. The long-term stability of the PMDT and PMTT membranes was investigated over 120 h under constant conditions of 0.05 bar and 50 °C for an aqueous solution of 95% n-butanol and 5% water. Although the stability performance of PMTT was constant for about 80 h, the permeation performance of PMDT reduced smoothly after this time. The long-term stability of PMDT was significantly constant for about 60 h. Therefore, the PMTT membrane in comparison with the PMDT sample exhibited more stability owing to its high density microstructure and PMDT with its low density microstructure showed lower stability.

Fig. 6 Long-term stability of PMDT and PMTT membranes at 50 °C over 120 h.

Conclusions

Two different membrane fabrication techniques were applied, namely dry-thermal (PMDT) and thermal treatment (PMTT) methods. These fabrication techniques enabled the preparation of low and high density membranes. The hybrid membranes were successfully fabricated by the incorporation of synthesized GO and experienced significantly lower contact angles in comparison with the nano plate GO membrane. We investigated the effect of the different fabrication techniques on the experimental performance on the n-butanol dehydration solution by observing the membrane morphology. The results indicated that PMDT and PMTT membranes displayed exceptional pervaporation performance in the constant concentration of 95% n-butanol and 5% water at different temperatures of 30, 40, and 50 °C. These membranes showed permeation fluxes of 1100.26 and 419.36 kg m⁻² h⁻¹ and separation factors of 87.44 and 96.97 at 30 °C for PMDT and PMTT, respectively. The permeation flux of the PMDT membrane decreased with increasing temperature, while the permeation performance increased for PMTT. Water permeation is reduced with increasing temperature due to the hydrogen bonding between the O–H group of water and the n-butanol molecules. This interaction led to the increase of the water molecule or water/alcohol molecule size, hence the water molecules cannot pass through the low density membranes. This fluctuation of the permeation flux of different pervaporation performances can be attributed to the morphology of the low and high density membranes.

References

1. Y. Huang, R. W. Baker and L. M. Vane, Ind. Eng. Chem. Res., 2010, 49, 3760–3768.
2. J. Neel, P. Aptel and R. Clement, Desalination, 1985, 53, 297–326.
3. W. Kujawski, Pol. J. Environ. Stud., 2000, 9, 13–26.
4. M. Mulder, Basic principles of membrane technology, Springer Science & Business Media, 1996, pp. 22–24.
5. Y. Huang, J. Ly, D. Nguyen and R. W. Baker, Ind. Eng. Chem. Res., 2010, 49, 12067–12073.
6. G. Li, L. Shi, G. Zeng, Y. Zhang and Y. Sun, RSC Adv., 2014, 4, 52012–52015.
7. L. M. Vane, V. V. Namboodiri and T. C. Bowen, J. Membr. Sci., 2008, 308, 230–241.
8. S. Sridhar, D. Ganga, B. Smitha and M. Ramakrishna, Sep. Sci. Technol., 2005, 40, 2889–2908.
9. C. H. Tsou, Q. F. An, S. C. Lo, M. De Guzman, W. S. Hung, C. C. Hu and J. Y. Lai, J. Membr. Sci., 2015, 477, 93–100.
10. X. Chen, G. Liu and Y. F. H. Zhang, Chin. J. Chem. Eng., 2015, 23, 1102–1109.
11. G. Liu, W. Wei and W. Jin, ACS Sustainable Chem. Eng., 2013, 2, 546–560.
12. F. Lipnizki, W. Robert and P. T. Kiong, J. Membr. Sci., 1999, 153, 183–210.
13. E. A. Fouad and X. Feng, J. Membr. Sci., 2008, 323, 428–435.
14. W. Van Hecke, T. Hofmann and H. De Wever, Bioresour. Technol., 2013, 129, 421–429.
15. N. Qureshi and T. C. Ezeji, Biofuels, Bioprod. Biorefin., 2008, 2, 319–330.
16. S. Guimin, Pervaporation Dehydration of Biofuels and Solvents, PhD thesis, National University of Singapore, 2013.
17. A. C. Hansen and D. C. Kyritsis, Biomass to biofuels: strategies for global industries, 2010, pp. 1–26.
18. K. Schmidt-Rohr, J. Chem. Educ., 2015, 92, 2094–2099.
19. Y. Dahman, Ferment. Technol., 2012, 1, e111.
20. T. Khosravi, S. Mosleh, O. Bakhtiari and T. Mohammadi, Chem. Eng. Res. Des., 2012, 90, 2353–2363.
21. B. Bolto, M. Hoang and Z. Xie, Chem. Eng. Process., 2011, 50, 227–235.
22. D. Rui, Z. Haoqin, L. Yifan, W. Jingtao, S. Benbing, M. Heng, D. Jingchuan and L. Jindun, J. Membr. Sci., 2015, 138, 227–238.
23. Y. Lou, G. Liu, S. Liu, J. Shen and W. Jin, Appl. Surf. Sci., 2014, 307, 631–637.
24. W. S. Hung, C. H. Tsou, M. D. Guzman, Q. F. An, Y. L. Liu, Y. M. Zhang, C. C. Hu, K. R. Lee and J. Y. Lai, Chem. Mater., 2014, 9, 2983–2990.
25. J. Ling, W. Zhai, W. Feng, B. Shen, J. Zhang and W. G. Zheng, ACS Appl. Mater. Interfaces, 2013, 5(7), 2677–2684.
26. E. J. Jeon, A. S. Kim and Y. T. Lee, Desalin. Water Treat., 2012, 48, 17–26.
27. D. Oh, S. Lee and Y. Lee, Desalin. Water Treat., 2013, 51, 5362–5370.
28. S. Amnuaypanich, J. Patthana and P. Phinyocheep, Chem. Eng. Sci., 2009, 64, 4908–4918.
29. R. Veerapur, M. B. Patil, K. B. Gudasi and T. M. Aminabhavi, Sep. Purif. Technol., 2008, 58, 377–385.
30. S. Amnuaypanich, T. Naowanon, W. Wongthep and P. Phinyocheep, J. Appl. Polym. Sci., 2012, 124, E319–E329.
31. Y. Zhu, S. Murali, W. Cai, X. Li, J. W. Suk, J. R. Potts and R. S. Ruoff, Graphene and graphene oxide: synthesis, properties, and applications, Adv. Mater., 2010, 22, 3906–3924.
32. A. M. Alamaria, M. G. M. Nawawi and Z. Zamrud, Arabian J. Chem., 2015, 9 DOI:10.1016/j.arabjc.2014.12.019.
33. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2010, 39, 228–240.
34. F. Kim, L. J. Cote and J. Huang, Adv. Mater., 2010, 22, 1954–1958.
35. Y. Zhu, S. Murali, W. Cai, X. Li, J. W. Suk, J. R. Potts and R. S. Ruoff, Adv. Mater., 2010, 22, 3906–3924.
36. G. Wang, S. Xiaoping, Y. Jane and P. Jinsoo, Carbon, 2009, 47, 2049–2053.
37. R. Nair, H. A. Wu, P. N. Jayaram, I. V. Grigorieva and A. K. Geim, Science, 2012, 335, 442–444.
38. Y. P. Tang, D. R. Paul and T. S. Chung, J. Membr. Sci., 2014, 458, 199–208.
39. E. Mahmoudi, L. Y. Ng, B. A. Abbad and A. W. Mohammad, Chem. Eng. J., 2015, 277, 1–10.
40. W. Xiaoli, H. Lan, Z. Jiakui, Z. Xiang, W. Jingtao and L. Jindun, J. Membr. Sci., 2016, 515, 175–188.
41. Y. Wang, M. Gruender and T. S. Chung, J. Membr. Sci., 2010, 363, 149–159.
42. J. Wijmans and R. Baker, J. Membr. Sci., 1995, 107, 1–21.
43. X. Fan, W. Peng, Y. Li, X. Li, S. Wang, G. Zhang and F. Zhang, Adv. Mater., 2008, 20, 4490–4493.
44. D. Li, M. B. Mueller, S. Gilje, R. B. Kaner and G. G. Wallace, Nat. Nanotechnol., 2008, 3, 101–105.
45. L. J. Cote, R. Cruz-Silva and J. Huang, J. Am. Chem. Soc., 2009, 131, 11027–11032.
46. M. N. Hyder, R. Y. M. Huang and P. Chen, J. Membr. Sci., 2008, 318, 387–396.
47. M. E. Uddina, R. K. Layeka, N. H. Kima, D. Huib and J. H. Lee, Composites, Part B, 2015, 80, 238–245.
48. J. L. Yan, C. P. Liu, Z. Wang, W. Xing and M. X. Ding, Polymer, 2007, 48, 6210–6214.
49. S. Zinadini, A. A. Zinatizadeh, M. Rahimi, V. Vatanpour and H. Zangeneh, J. Membr. Sci., 2014, 453, 292–301.

---