DOI: 10.1002/chem.201504836
Communication
& Materials Science
Chemically Stable Covalent Organic Framework (COF)Polybenzimidazole Hybrid Membranes: Enhanced Gas Separation through Pore Modulation Bishnu P. Biswal,[a] Harshal D. Chaudhari,[b] Rahul Banerjee,*[a] and Ulhas K. Kharul*[b] Abstract: Highly flexible, TpPa-1@PBI-BuI and TpBD@PBIBuI hybrid membranes based on chemically stable covalent organic frameworks (COFs) could be obtained with the polymer. The loading obtained was substantially higher (50 %) than generally observed with MOFs. These hybrid membranes show an exciting enhancement in permeability (about sevenfold) with appreciable separation factors for CO2/N2 and CO2/CH4. Further, we found that with COF pore modulation, the gas permeability can be systematically enhanced.
The fabrication of composite membranes for molecular separation has gathered a great deal of interest in recent years.[1] Among known membrane materials, polymeric ones are always preferred for various separation applications due to their easy processability, mechanical stability and scale-up opportunity.[2] However, the applications of polymeric membranes are limited by permeability–selectivity trade-off, plasticization, physical aging and high temperature withstand capacity, which proves to be a major concern and needs urgent attention.[2c] To overcome these shortcomings, researchers are developing various methodologies of incorporating porous materials such as zeolite, carbon nanotubes (CNTs), carbon molecular sieve (CMS), metal–organic frameworks (MOFs), porous aromatic frameworks (PAFs) and porous organic cages as fillers inside the polymer matrix to create synergistic enhancements in the performance of resulting composite membranes.[3] Although a significant advancement in MOF-based molecular sieve membranes has been documented in the literature, such efforts often face compatibility issues with the polymers, which leads to cracks and defects in the resulting composite membranes.[4] [a] B. P. Biswal, Prof. Dr. R. Banerjee Academy of Scientific and Innovative Research (AcSIR) Physical/Materials Chemistry Division, CSIR-National Chemical Laboratory Dr. Homi Bhabha Road, Pune-411008 (India) E-mail: [email protected] [b] H. D. Chaudhari, Prof. Dr. U. K. Kharul Academy of Scientific and Innovative Research (AcSIR) Polymer Science and Engineering Division CSIR-National Chemical Laboratory Dr. Homi Bhabha Road, Pune-411008 (India) E-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201504836. Chem. Eur. J. 2016, 22, 4695 – 4699
This motivated us to use fully organic materials that are crystalline, porous, and possess well-defined nanochannels and high chemical stability (long lifetimes) as an efficient filler. Due to their fully organic nature, they are anticipated to offer excellent compatibility with the polymer matrix used for membrane preparation. Covalent organic frameworks (COFs) are crystalline, porous materials constructed by strong covalent bonds linked with lightweight elements (H, B, C, Si, N, O).[5] These materials have attracted interest in areas like adsorption/storage, chemical sensors and catalysis due to their highly ordered and low-density framework, with an opportunity to incorporate diverse functional groups at a molecular level.[6] Despite such potentials, COF-based polymer hybrid membranes are yet to be explored for molecular separation.[7] This could be due to the lack of chemical stability in the attempted COFs under operational conditions, which prevents their use in making membranes for gas/liquid separation.[8] We have recently addressed the issue of chemical stability of COFs by introducing ketoenol tautomerism within their framework, which imparts high water, acid, and base stability.[9] This outstanding stability feature, along with well-defined and tunable porous nanochannels in these COFs led us to make hybrid membranes for selective molecular transport for gas/liquid separation. Herein, we report for the first time, usage of chemically stable isoreticular COFs (TpPa-1 and TpBD, pore aperture of 18 and 24 æ, respectively) as active phase to be incorporated within the polymer (PBI-BuI) matrix for making self-supported TpPa-1@PBI-BuI and TpBD@PBI-BuI hybrid membranes. Considerably higher loading of COF in polymer matrix could be achieved (50 %) than that generally observed for common MOFs ( 30 %). These COF@Polymer hybrid membranes possess exceptionally high chemical stability, high flexibility and thus, can be easily processed. Our strategy was to create intermolecular interactions between H-bonded benzimidazole groups of PBI with COFs to improve the filler loading and thereby enhance overall permeability of the composite matrix. Six hybrid membranes, namely, TpPa-1(20)@PBI-BuI, TpPa-1(40)@PBI-BuI, TpPa1(50)@PBI-BuI, TpBD(20)@PBI-BuI, TpBD(40)@PBI-BuI and TpBD(50)@PBI-BuI were prepared with a sequential increase of COF content in the polymer, PBI-BuI. We selected the substituted polybenzimidazole (PBI-BuI) as it exhibits good permeability by itself and compatibility with COFs. This way, both partners contain organic backbone and H-bonding sites (thus compatibility) and thermochemical stability, which are requisites for membrane usability for real-life applications (e.g., separations
4695
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication involving H2 and CO2 at elevated temperatures). The so-obtained COF@PBI-BuI hybrid membranes were evaluated for gas permeation and were found to offer substantially elevated (about sevenfold) permeability (H2, N2, CH4 and CO2), while maintaining appreciable selectivities with respect to the pristine PBI-BuI membrane. PBI-BuI was synthesized following our earlier report[10d] and vacuum dried at 100 8C for 24 h prior to use. TpPa-1 and TpBD were synthesized by using our previous protocol and were freshly characterized.[9b] A solution-casting method using DMAc (N,N-Dimethylacetamide) as the solvent was employed for the fabrication of highly flexible COF(n)@PBI-BuI hybrid membranes with varying COF content (where, n = 20, 40 or 50 wt % of TpPa-1 and TpBD; Figure 1 and Section S1 in the Supporting Information). In these hybrid membranes, the COF loading could be successfully achieved to 50 %. Beyond this composition, defects in the membrane were observed. Importantly, all prepared COF–polymer hybrids are quite flexible even at higher (50 %) COF loading, unlike their MOF membrane counterparts where only 30 % ZIF-8 loading was possible.[10c] The wide-angle X-ray diffraction (WAXD) patterns of the assynthesized TpPa-1 and TpBD matched well with their corresponding simulated patterns. The pristine PBI-BuI membrane showed one major broad WAXD peak in the 2q range of 158– 308, which indicates its amorphous nature. The TpPa-1(n)@PBI-
BuI and TpBD(n)@PBI-BuI membranes were analyzed by WAXD to ensure the phase purity of TpPa-1 and TpBD embedded inside the matrix (Figure S2 in the Supporting Information). In the WAXD pattern of hybrid membranes, a crystalline peak at 4.78 and 3.48 (2q), corresponds to 100 planes of TpPa-1 and TpBD and confirms their stability in the formed hybrid membranes. Also, the characteristic amorphous hump of host PBIBuI was observed. The FT-IR spectrum of pristine PBI-BuI-HF was characterized by absorption in the range of 1430– 1650 cm¢1 for benzimidazole repeat units. The broad band at approximately 3145 cm¢1 was ascribed to the N¢H···N hydrogen bond and the peak at 2862 cm¢1 is due to the presence of the tert-butyl group of PBI-BuI.[10d] However, for TpPa-1@PBIBuI and TpBD@PBI-BuI hybrid membranes, few major bands of TpPa-1 and TpBD are found to be merged with the pristine PBI-BuI. The appearance of a peak at approximately 1578 cm¢1 is due to the exocyclic C=C bond of Tp. The peaks at 1445 cm¢1 [C=C(Ar)] and 1256 cm¢1 (C¢N), are due to the aromatic C=C and C¢N bond in the keto-enamine form of TpPa1 and TpBD (Figure S3 in the Supporting Information). The scanning electron microscope (SEM) images indicated that TpPa-1 and TpBD have a spherical flower-like morphology with an average size of 5–7 mm. The SEM cross section of the TpBD@PBI-BuI hybrid membranes (Figure 2) confirms the distribution of spherical COF particles throughout the membrane
Figure 1. A) Schematic representations of the synthesis of COFs and their packing models indicating the pore aperture and stacking distances. B) Overview of the solution-casting method for COF@PBI-BuI hybrid membrane fabrication. C) Digital photographs showing the flexibility of TpPa-1 and TpBD(50)@PBI-BuI hybrid membranes. Chem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
4696
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication matrix. Further, it has been noticed that the average thicknesses of the representative hybrid membranes are approximately 47 to 80 mm. In both cases the membrane cross section and the surface did not show any visible cracks or tears at the COF–polymer interface (Figure 2, and Figures S6 and S7 in the Supporting Information). Further, the COF particles are firmly bound with the PBI-BuI backbone, which does not allow the COFs to leach out from the polymer matrix. This observation indicated a very good compatibility and adhesion between the polymer and COFs, as both contain pure organic backbone with H-bonding sites.[11] Thermogravimetric analysis (TGA) was performed to understand the thermal behavior of these hybrid membranes. TGA profiles indicated that the COFs (TpPa-1 and TpBD) have guest-free pores, and no visible weight loss was observed until 350 8C, indicating their excellent thermal stability. However, a gradual weight loss of approximately 65 % (TpPa-1) and 90 % (TpPa-1) was observed up to 800 8C. PBI-BuI showed thermal stability at 525 8C, which is consistent with the literature reports.[10b] All these hybrid membranes showed a thermal stability up to approximately 400 8C, which lies between the thermal stability of PBI-BuI and COFs (Figure S5 in the Supporting Information). This higher stability of the hybrid membranes than that of COFs is a sign of attractive interactions (and thus good compatibility) between COFs and PBI-BuI. Moreover, mechanical properties analysis of COF(n)@PBI-BuI hybrid membranes showed a decrease of tensile strength and modulus with an increase in COF loading compared to the pristine PBI-BuI membrane (Figure S6 in the Supporting Information). This indicates that at higher COF loading (50 %) the continuity of the polymer phase is decreased. Polybenzimidazoles (PBIs) are well known for many exciting applications owing to their excellent thermochemical stability and outstanding mechanical properties at high temperature. Based on the considerations above, our motivation is to demonstrate elevation of gas permeance properties of PBI-BuI by introducing a 2D organic porous crystalline framework material in it. The beauty of these crystalline 2D materials is the chemical stability, tunable porosity and an opportunity for pore engineering. Further, it is expected that by judicious selection of COFs with different pore aperture, the molecular sieving properties can be achieved that are essential for various gas/liquid separations. We performed gas permeance studies using H2, N2, CH4 and CO2 at 35 8C and 20 atm upstream pressure. Such a high-pressure endurance capacity for composite membranes composed of COFs with high filler loading ( 50 %) has not been reported to date. The permeation data presented here is the average numbers of three samples (3.8 cm active membrane diameter). Figure 3 a and b, show the H2 and CO2 permeability of the PBI-BuI-based hybrid, which are found to increase almost linearly with the amount of COF loading. A three times elevation in H2 permeability from 6.2 Barrer (for pristine PBIBuI) to 18.8 Barrer was observed for TpPa-1(40)@PBI-BuI hybrid. This is associated with an increase in H2/CH4 selectivity from 155 to 165.5 and H2/N2 selectivity from 69 to 79. Although CO2/N2 selectivities (Figure 3 b) are slightly decreased, CO2/CH4 selectivity remained appreciable (46.3) even with 40 % TpPa-1 loading. This is found to be higher than the selectivity Chem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
Figure 2. SEM images showing cross sections of: a) TpBD, b) TpBD(20)@PBIBuI, c) TpBD(40)@PBI-BuI, and d) TpBD(50)@PBI-BuI membranes; and e– h) their respective zoomed views. Scale bars represent 10 mm (a–d) and 5 mm (e–h).
obtain for commonly used gas separation membrane materials such as Matrimid, polysulfone (PSF) and polycarbonate (PC; a(CO2/CH4) = 36, 22 and 19, respectively).[12] Further, we compared our results with some common membrane materials (PSF, Matrimid, PC, PPO) with reference to Robeson’s upper bound[2c] (Figure 3 c and d). Figure 3 a and b depict how the TpBD(50)@PBI-BuI hybrid outperforms the aforementioned materials with respect to CO2 permeability and that they are comparable in CO2/CH4 selectivity, except for PPO, where the selectivity is lower, but the permeability is higher. We also compared our gas permeation results with the previously reported ZIF-8@PBI-BuI-based hybrid membranes. It was found that the TpBD(50)@PBI-BuI-based membranes show better CO2 (14.8) and CH4 (0.3) permeability, with slightly higher CO2/CH4 selectivity (48.7) as compared to Z30@PBI-BuI (PCO2 : 5.23, PCH4 : 0.12 and CO2/CH4 selectivity was 43.6).[10c] The effect of pore modulation, aperture 18 æ (TpPa-1) to 24 æ (TpBD) of COF in gas permeation was prominently observed. Almost sevenfold elevation in H2 permeability compared to the pristine case was achieved with 50 % TpBD loading into PBI-BuI (Figure 3 a). This was coupled with comparable H2/N2 (from 69 to 66) and a slight decrease in H2/CH4 (155.5 to 139.7) selectivity than that of unloaded PBI-BuI. Thus, these selectivity values are still very well comparable with that of com-
4697
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
Figure 3. Variation in: a) H2 permeability and its selectivity over N2 and CH4 ; b) CO2 permeability and its selectivity over N2 and CH4 with respect to COF loading in TpPa-1@PBI-BuI hybrid membranes. c d) The single gas CO2 permeability and ideal CO2/CH4 and CO2/N2 selectivity of TpPa-1@PBI-BuI (black, filled squares) and TpBD@PBI-BuI (red, filled circles) hybrid membranes are plotted on Robeson’s upper bound.
monly used commercial gas separation membrane materials like PSF. The H2 permeability for PSF is reported to be 14 Barrer, which is three times lower than the H2 permeability (42.5 Barrer) of TpBD(50)@PBI-BuI (Table S1 in the Supporting Information). An increase in CO2 permeability from 3.2 Barrer (for pristine PBI-BuI) to 13.1 and 14.8 Barrer for TpPa1(50)@PBI-BuI and TpBD(50)@PBI-BuI hybrid membrane was generous (Figure 3 b). One of the peculiarities seen in the case of COF@PBI-BuI-based hybrid membranes was that the increase in permeability from 20 to 50 % of COF loading was linear. When the elevation in permeability of TpPa-1-loaded PBI-BuI is compared with TpBD-loaded PBI-BuI, the effect of pore modulation on elevating permeability, is prominently evident. Improved gas permeation while maintaining the base polymer (PBI-BuI) selectivity in the case of COF@PBI-BuI hybrid membranes highlights how the benefits of COF pores can be better achieved for lowering the diffusion resistance and thus elevating gas permeability. However the selectivity is probably still governed by the polymer matrix, that is, PBI-BuI. This could be made possible by avoiding agglomeration or such defect inChem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
hibition. As the COF pores ( 15 æ for TpPa-1 and 18 æ for TpBD) are much larger than the kinetic diameters of gases [2.89 æ (H2), 3.64 æ (N2), 3.3 æ (CO2) and 3.8 æ (CH4)], molecular sieving of gases (enhancement of selectivity) is not really expected from COFs. Notably, in some of the MOF-based MMMs (Mixed Matrix Membranes), high enhancement in selectivity has been evidenced due to narrow pore aperture (3–4 æ) of MOFs. In the present case, the COFs (TpBD) contribute to a significant increase in permeability of PBI-BuI, with small deviations in selectivity. It can thus be further expected that if the COFs with even higher pore aperture are chosen, elevation in permeability could be more pronounced. In conclusion, we have showcased, for the first time, a methodology to fabricate a new class of self-supported COF@polymer hybrid membranes. These self-supported membranes are highly flexible, reproducible and display a high degree of thermal and chemical stabilities; they show potential for a wider range of applications in gas/liquid separation. As proof-of-concept, we performed gas-permeation analysis, which revealed an almost sevenfold elevation in permeability of gases (H2, N2, CH4 and CO2) compared with the pristine membrane with 50 %
4698
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication TpBD loading and appreciable CO2/N2 (23) and CO2/CH4 (48.7) selectivity. Interestingly, the effect of isoreticulation in COF towards elevating the gas permeability was prominently seen in these COF@PBI-BuI-based hybrid membranes. Overall, the findings demonstrate a versatile approach of preparing chemically stable functional COF-based hybrid membranes that could significantly contribute to the advancement of such materials in the near future. [4]
Experimental Section
[5]
TpPa-1 and TpBD were synthesized using our standard reported solvothermal syntheses method.[9] A solution casting method using DMAc as a solvent was employed for the fabrication of COF(n)@PBI-BuI hybrid membranes by varying COF content (where n = 20, 40 and 50 % by weight of TpPa-1 and TpBD). The detailed materials and experimental methods are described in the Supporting Information.
[6]
[7]
Acknowledgements
[8]
B.P.B. and H.D.C. acknowledge the UGC and CSIR (New Delhi, India) for SRF. R.B. and U.K.K. acknowledge the CSIR (CSC0122 and CSC0102), DST Indo-Singapore Project (INT/SIN/P-05) and DST Nano-mission Project (SR/NM/NS-1179/2012G) for funding.
[9]
[10]
Keywords: covalent organic frameworks · gas separation · membranes · microporous materials · pore modulation
[11]
[1] a) S. Keskin, T. M. van Heest, D. S. Sholl, ChemSusChem 2010, 3, 879; b) S. Basu, A. Cano-Odena, I. F. J. Vankelecom, J. Membr. Sci. 2010, 362, 478; c) H. B. T. Jeazet, C. Staudt, C. Janiak, Chem. Commun. 2012, 48, 2140; d) D. Nagaraju, D. G. Bhagat, R. Banerjee, U. K. Kharul, J. Mater. Chem. A 2013, 1, 8828; e) A. J. Brown, N. A. Brunelli, K. Eum, F. Rashidi, J. R. Johnson, W. J. Koros, C. W. Jones, S. Nair, Science 2014, 345, 72. [2] a) D. L. Gin, R. D. Noble, Science 2011, 332, 674; b) N. Du, H. B. Park, G. P. Robertson, M. M. Dal-Cin, T. Visser, L. Scoles, M. D. Guiver, Nat. Mater. 2011, 10, 372; c) L. M. Robeson, J. Membr. Sci. 2008, 320, 390. [3] a) B. A. Al-Maythalony, O. Shekhah, R. Swaidan, Y. Belmabkhout, I. Pinnau, M. Eddaoudi, J. Am. Chem. Soc. 2015, 137, 1754; b) S. Kim, L.
Chem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
[12]
Chen, J. K. Johnson, E. Marand, J. Membr. Sci. 2007, 294, 147; c) D. Q. Vu, W. J. Koros, S. J. Miller, J. Membr. Sci. 2003, 211, 311; d) C. H. Lau, P. T. Nguyen, M. R. Hill, A. W. Thornton, K. Konstas, C. M. Doherty, R. J. Mulder, L. Bourgeois, A. C. Y. Liu, D. J. Sprouster, J. P. Sullivan, T. J. Bastow, A. J. Hill, D. L. Gin, R. D. Noble; Angew. Chem. Int. Ed. 2014, 53, 5322; Angew. Chem. 2014, 126, 5426; Angew. Chem. Int. Ed. 2014, 53, 5322; Angew. Chem. 2014, 126, 5426; e) A. F. Bushell, P. M. Budd, M. P. Attfield, J. T. A. Jones, T. Hasell, A. I. Cooper, P. Bernardo, F. Bazzarelli, G. Clarizia, J. C. Jansen, Angew. Chem. Int. Ed. 2013, 52, 1253; Angew. Chem. 2013, 125, 1291. B. Zornoza, C. Tellez, J. Coronas, J. Gascon, F. Kapteijn, Microporous Mesoporous Mater. 2013, 166, 67. a) A. P. Cút¦, A. I. Benin, N. W. Ockwig, A. J. Matzger, M. O’Keeffe, O. M. Yaghi, Science 2005, 310, 1166; b) X. Feng, X. Dinga, D. Jiang, Chem. Soc. Rev. 2012, 41, 6010. a) J. F. Dienstmaier, D. D. Medina, M. Dogru, P. Knochel, T. Bein, W. M. Heckl, M. Lackinger, ACS Nano 2012, 6, 7234; b) S. Y. Ding, J. Gao, Q. Wang, Y. Zhang, W. G. Song, C. Y. Su, W. Wang, J. Am. Chem. Soc. 2011, 133, 19816; c) S. Wan, J. Guo, J. Kim, H. Ihee, D. Jiang, Angew. Chem. Int. Ed. 2009, 48, 5439; Angew. Chem. 2009, 121, 5547. a) D. D. Hao, J. N. Zhang, H. Lu, W. G. Leng, R. L. Ge, X. N. Dai, Y. Gao, Chem. Commun. 2014, 50, 1462; b) H. Lu, C. Wang, J. Chen, R. Ge, W. Leng, B. Dong, J. Huang, Y. Gao, Chem. Commun. 2015, 51, 15562. a) L. M. Lanni, R. W. Tilford, M. Bharathy, J. J. Lavigne, J. Am. Chem. Soc. 2011, 133, 13975; b) E. Spitler, M. Giovino, S. White, W. Dichtel, Chem. Sci. 2011, 2, 1588. a) S. Kandambeth, A. Mallick, B. Lukose, V. M. Mane, T. Heine, R. Banerjee, J. Am. Chem. Soc. 2012, 134, 19524; b) S. Chandra, S. Kandambeth, B. P. Biswal, B. Lukose, S. M. Kunjir, M. Chaudhary, R. Babarao, T. Heine, R. Banerjee, J. Am. Chem. Soc. 2013, 135, 17853. a) T. Yang, Y. Xiao, T. S. Chung, Energy Environ. Sci. 2011, 4, 4171; b) S. C. Kumbharkar, K. Li, J. Membr. Sci. 2012, 415, 793; c) A. Bhaskar, R. Banerjee, U. K. Kharul, J. Mater. Chem. A 2014, 2, 12962; d) S. C. Kumbharkar, P. B. Karadkar, U. K. Kharul, J. Membr. Sci. 2006, 286, 161. a) G. R. Desiraju, Acc. Chem. Res. 2002, 35, 565; b) G. R. Desiraju, T. Steiner, The Weak Hydrogen Bond in Structural Chemistry and Biology, Oxford University Press, Oxford, 1999. a) W. F. Yong, F. Y. Li, Y. C. Xiao, P. Li, K. P. Pramod, Y. W. Tong, T. S. Chung, J. Membr. Sci. 2012, 47, 407; b) J. S. McHattie, W. J. Koros, D. R. Paul, Polymer 1992, 33, 1701; c) M. W. Hellums, W. J. Koros, G. R. Husk, D. R. Paul, J. Membr. Sci. 1989, 46, 93.
Received: December 2, 2015 Published online on February 18, 2016
4699
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
& Materials Science
Chemically Stable Covalent Organic Framework (COF)Polybenzimidazole Hybrid Membranes: Enhanced Gas Separation through Pore Modulation Bishnu P. Biswal,[a] Harshal D. Chaudhari,[b] Rahul Banerjee,*[a] and Ulhas K. Kharul*[b] Abstract: Highly flexible, TpPa-1@PBI-BuI and TpBD@PBIBuI hybrid membranes based on chemically stable covalent organic frameworks (COFs) could be obtained with the polymer. The loading obtained was substantially higher (50 %) than generally observed with MOFs. These hybrid membranes show an exciting enhancement in permeability (about sevenfold) with appreciable separation factors for CO2/N2 and CO2/CH4. Further, we found that with COF pore modulation, the gas permeability can be systematically enhanced.
The fabrication of composite membranes for molecular separation has gathered a great deal of interest in recent years.[1] Among known membrane materials, polymeric ones are always preferred for various separation applications due to their easy processability, mechanical stability and scale-up opportunity.[2] However, the applications of polymeric membranes are limited by permeability–selectivity trade-off, plasticization, physical aging and high temperature withstand capacity, which proves to be a major concern and needs urgent attention.[2c] To overcome these shortcomings, researchers are developing various methodologies of incorporating porous materials such as zeolite, carbon nanotubes (CNTs), carbon molecular sieve (CMS), metal–organic frameworks (MOFs), porous aromatic frameworks (PAFs) and porous organic cages as fillers inside the polymer matrix to create synergistic enhancements in the performance of resulting composite membranes.[3] Although a significant advancement in MOF-based molecular sieve membranes has been documented in the literature, such efforts often face compatibility issues with the polymers, which leads to cracks and defects in the resulting composite membranes.[4] [a] B. P. Biswal, Prof. Dr. R. Banerjee Academy of Scientific and Innovative Research (AcSIR) Physical/Materials Chemistry Division, CSIR-National Chemical Laboratory Dr. Homi Bhabha Road, Pune-411008 (India) E-mail: [email protected] [b] H. D. Chaudhari, Prof. Dr. U. K. Kharul Academy of Scientific and Innovative Research (AcSIR) Polymer Science and Engineering Division CSIR-National Chemical Laboratory Dr. Homi Bhabha Road, Pune-411008 (India) E-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201504836. Chem. Eur. J. 2016, 22, 4695 – 4699
This motivated us to use fully organic materials that are crystalline, porous, and possess well-defined nanochannels and high chemical stability (long lifetimes) as an efficient filler. Due to their fully organic nature, they are anticipated to offer excellent compatibility with the polymer matrix used for membrane preparation. Covalent organic frameworks (COFs) are crystalline, porous materials constructed by strong covalent bonds linked with lightweight elements (H, B, C, Si, N, O).[5] These materials have attracted interest in areas like adsorption/storage, chemical sensors and catalysis due to their highly ordered and low-density framework, with an opportunity to incorporate diverse functional groups at a molecular level.[6] Despite such potentials, COF-based polymer hybrid membranes are yet to be explored for molecular separation.[7] This could be due to the lack of chemical stability in the attempted COFs under operational conditions, which prevents their use in making membranes for gas/liquid separation.[8] We have recently addressed the issue of chemical stability of COFs by introducing ketoenol tautomerism within their framework, which imparts high water, acid, and base stability.[9] This outstanding stability feature, along with well-defined and tunable porous nanochannels in these COFs led us to make hybrid membranes for selective molecular transport for gas/liquid separation. Herein, we report for the first time, usage of chemically stable isoreticular COFs (TpPa-1 and TpBD, pore aperture of 18 and 24 æ, respectively) as active phase to be incorporated within the polymer (PBI-BuI) matrix for making self-supported TpPa-1@PBI-BuI and TpBD@PBI-BuI hybrid membranes. Considerably higher loading of COF in polymer matrix could be achieved (50 %) than that generally observed for common MOFs ( 30 %). These COF@Polymer hybrid membranes possess exceptionally high chemical stability, high flexibility and thus, can be easily processed. Our strategy was to create intermolecular interactions between H-bonded benzimidazole groups of PBI with COFs to improve the filler loading and thereby enhance overall permeability of the composite matrix. Six hybrid membranes, namely, TpPa-1(20)@PBI-BuI, TpPa-1(40)@PBI-BuI, TpPa1(50)@PBI-BuI, TpBD(20)@PBI-BuI, TpBD(40)@PBI-BuI and TpBD(50)@PBI-BuI were prepared with a sequential increase of COF content in the polymer, PBI-BuI. We selected the substituted polybenzimidazole (PBI-BuI) as it exhibits good permeability by itself and compatibility with COFs. This way, both partners contain organic backbone and H-bonding sites (thus compatibility) and thermochemical stability, which are requisites for membrane usability for real-life applications (e.g., separations
4695
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication involving H2 and CO2 at elevated temperatures). The so-obtained COF@PBI-BuI hybrid membranes were evaluated for gas permeation and were found to offer substantially elevated (about sevenfold) permeability (H2, N2, CH4 and CO2), while maintaining appreciable selectivities with respect to the pristine PBI-BuI membrane. PBI-BuI was synthesized following our earlier report[10d] and vacuum dried at 100 8C for 24 h prior to use. TpPa-1 and TpBD were synthesized by using our previous protocol and were freshly characterized.[9b] A solution-casting method using DMAc (N,N-Dimethylacetamide) as the solvent was employed for the fabrication of highly flexible COF(n)@PBI-BuI hybrid membranes with varying COF content (where, n = 20, 40 or 50 wt % of TpPa-1 and TpBD; Figure 1 and Section S1 in the Supporting Information). In these hybrid membranes, the COF loading could be successfully achieved to 50 %. Beyond this composition, defects in the membrane were observed. Importantly, all prepared COF–polymer hybrids are quite flexible even at higher (50 %) COF loading, unlike their MOF membrane counterparts where only 30 % ZIF-8 loading was possible.[10c] The wide-angle X-ray diffraction (WAXD) patterns of the assynthesized TpPa-1 and TpBD matched well with their corresponding simulated patterns. The pristine PBI-BuI membrane showed one major broad WAXD peak in the 2q range of 158– 308, which indicates its amorphous nature. The TpPa-1(n)@PBI-
BuI and TpBD(n)@PBI-BuI membranes were analyzed by WAXD to ensure the phase purity of TpPa-1 and TpBD embedded inside the matrix (Figure S2 in the Supporting Information). In the WAXD pattern of hybrid membranes, a crystalline peak at 4.78 and 3.48 (2q), corresponds to 100 planes of TpPa-1 and TpBD and confirms their stability in the formed hybrid membranes. Also, the characteristic amorphous hump of host PBIBuI was observed. The FT-IR spectrum of pristine PBI-BuI-HF was characterized by absorption in the range of 1430– 1650 cm¢1 for benzimidazole repeat units. The broad band at approximately 3145 cm¢1 was ascribed to the N¢H···N hydrogen bond and the peak at 2862 cm¢1 is due to the presence of the tert-butyl group of PBI-BuI.[10d] However, for TpPa-1@PBIBuI and TpBD@PBI-BuI hybrid membranes, few major bands of TpPa-1 and TpBD are found to be merged with the pristine PBI-BuI. The appearance of a peak at approximately 1578 cm¢1 is due to the exocyclic C=C bond of Tp. The peaks at 1445 cm¢1 [C=C(Ar)] and 1256 cm¢1 (C¢N), are due to the aromatic C=C and C¢N bond in the keto-enamine form of TpPa1 and TpBD (Figure S3 in the Supporting Information). The scanning electron microscope (SEM) images indicated that TpPa-1 and TpBD have a spherical flower-like morphology with an average size of 5–7 mm. The SEM cross section of the TpBD@PBI-BuI hybrid membranes (Figure 2) confirms the distribution of spherical COF particles throughout the membrane
Figure 1. A) Schematic representations of the synthesis of COFs and their packing models indicating the pore aperture and stacking distances. B) Overview of the solution-casting method for COF@PBI-BuI hybrid membrane fabrication. C) Digital photographs showing the flexibility of TpPa-1 and TpBD(50)@PBI-BuI hybrid membranes. Chem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
4696
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication matrix. Further, it has been noticed that the average thicknesses of the representative hybrid membranes are approximately 47 to 80 mm. In both cases the membrane cross section and the surface did not show any visible cracks or tears at the COF–polymer interface (Figure 2, and Figures S6 and S7 in the Supporting Information). Further, the COF particles are firmly bound with the PBI-BuI backbone, which does not allow the COFs to leach out from the polymer matrix. This observation indicated a very good compatibility and adhesion between the polymer and COFs, as both contain pure organic backbone with H-bonding sites.[11] Thermogravimetric analysis (TGA) was performed to understand the thermal behavior of these hybrid membranes. TGA profiles indicated that the COFs (TpPa-1 and TpBD) have guest-free pores, and no visible weight loss was observed until 350 8C, indicating their excellent thermal stability. However, a gradual weight loss of approximately 65 % (TpPa-1) and 90 % (TpPa-1) was observed up to 800 8C. PBI-BuI showed thermal stability at 525 8C, which is consistent with the literature reports.[10b] All these hybrid membranes showed a thermal stability up to approximately 400 8C, which lies between the thermal stability of PBI-BuI and COFs (Figure S5 in the Supporting Information). This higher stability of the hybrid membranes than that of COFs is a sign of attractive interactions (and thus good compatibility) between COFs and PBI-BuI. Moreover, mechanical properties analysis of COF(n)@PBI-BuI hybrid membranes showed a decrease of tensile strength and modulus with an increase in COF loading compared to the pristine PBI-BuI membrane (Figure S6 in the Supporting Information). This indicates that at higher COF loading (50 %) the continuity of the polymer phase is decreased. Polybenzimidazoles (PBIs) are well known for many exciting applications owing to their excellent thermochemical stability and outstanding mechanical properties at high temperature. Based on the considerations above, our motivation is to demonstrate elevation of gas permeance properties of PBI-BuI by introducing a 2D organic porous crystalline framework material in it. The beauty of these crystalline 2D materials is the chemical stability, tunable porosity and an opportunity for pore engineering. Further, it is expected that by judicious selection of COFs with different pore aperture, the molecular sieving properties can be achieved that are essential for various gas/liquid separations. We performed gas permeance studies using H2, N2, CH4 and CO2 at 35 8C and 20 atm upstream pressure. Such a high-pressure endurance capacity for composite membranes composed of COFs with high filler loading ( 50 %) has not been reported to date. The permeation data presented here is the average numbers of three samples (3.8 cm active membrane diameter). Figure 3 a and b, show the H2 and CO2 permeability of the PBI-BuI-based hybrid, which are found to increase almost linearly with the amount of COF loading. A three times elevation in H2 permeability from 6.2 Barrer (for pristine PBIBuI) to 18.8 Barrer was observed for TpPa-1(40)@PBI-BuI hybrid. This is associated with an increase in H2/CH4 selectivity from 155 to 165.5 and H2/N2 selectivity from 69 to 79. Although CO2/N2 selectivities (Figure 3 b) are slightly decreased, CO2/CH4 selectivity remained appreciable (46.3) even with 40 % TpPa-1 loading. This is found to be higher than the selectivity Chem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
Figure 2. SEM images showing cross sections of: a) TpBD, b) TpBD(20)@PBIBuI, c) TpBD(40)@PBI-BuI, and d) TpBD(50)@PBI-BuI membranes; and e– h) their respective zoomed views. Scale bars represent 10 mm (a–d) and 5 mm (e–h).
obtain for commonly used gas separation membrane materials such as Matrimid, polysulfone (PSF) and polycarbonate (PC; a(CO2/CH4) = 36, 22 and 19, respectively).[12] Further, we compared our results with some common membrane materials (PSF, Matrimid, PC, PPO) with reference to Robeson’s upper bound[2c] (Figure 3 c and d). Figure 3 a and b depict how the TpBD(50)@PBI-BuI hybrid outperforms the aforementioned materials with respect to CO2 permeability and that they are comparable in CO2/CH4 selectivity, except for PPO, where the selectivity is lower, but the permeability is higher. We also compared our gas permeation results with the previously reported ZIF-8@PBI-BuI-based hybrid membranes. It was found that the TpBD(50)@PBI-BuI-based membranes show better CO2 (14.8) and CH4 (0.3) permeability, with slightly higher CO2/CH4 selectivity (48.7) as compared to Z30@PBI-BuI (PCO2 : 5.23, PCH4 : 0.12 and CO2/CH4 selectivity was 43.6).[10c] The effect of pore modulation, aperture 18 æ (TpPa-1) to 24 æ (TpBD) of COF in gas permeation was prominently observed. Almost sevenfold elevation in H2 permeability compared to the pristine case was achieved with 50 % TpBD loading into PBI-BuI (Figure 3 a). This was coupled with comparable H2/N2 (from 69 to 66) and a slight decrease in H2/CH4 (155.5 to 139.7) selectivity than that of unloaded PBI-BuI. Thus, these selectivity values are still very well comparable with that of com-
4697
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
Figure 3. Variation in: a) H2 permeability and its selectivity over N2 and CH4 ; b) CO2 permeability and its selectivity over N2 and CH4 with respect to COF loading in TpPa-1@PBI-BuI hybrid membranes. c d) The single gas CO2 permeability and ideal CO2/CH4 and CO2/N2 selectivity of TpPa-1@PBI-BuI (black, filled squares) and TpBD@PBI-BuI (red, filled circles) hybrid membranes are plotted on Robeson’s upper bound.
monly used commercial gas separation membrane materials like PSF. The H2 permeability for PSF is reported to be 14 Barrer, which is three times lower than the H2 permeability (42.5 Barrer) of TpBD(50)@PBI-BuI (Table S1 in the Supporting Information). An increase in CO2 permeability from 3.2 Barrer (for pristine PBI-BuI) to 13.1 and 14.8 Barrer for TpPa1(50)@PBI-BuI and TpBD(50)@PBI-BuI hybrid membrane was generous (Figure 3 b). One of the peculiarities seen in the case of COF@PBI-BuI-based hybrid membranes was that the increase in permeability from 20 to 50 % of COF loading was linear. When the elevation in permeability of TpPa-1-loaded PBI-BuI is compared with TpBD-loaded PBI-BuI, the effect of pore modulation on elevating permeability, is prominently evident. Improved gas permeation while maintaining the base polymer (PBI-BuI) selectivity in the case of COF@PBI-BuI hybrid membranes highlights how the benefits of COF pores can be better achieved for lowering the diffusion resistance and thus elevating gas permeability. However the selectivity is probably still governed by the polymer matrix, that is, PBI-BuI. This could be made possible by avoiding agglomeration or such defect inChem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
hibition. As the COF pores ( 15 æ for TpPa-1 and 18 æ for TpBD) are much larger than the kinetic diameters of gases [2.89 æ (H2), 3.64 æ (N2), 3.3 æ (CO2) and 3.8 æ (CH4)], molecular sieving of gases (enhancement of selectivity) is not really expected from COFs. Notably, in some of the MOF-based MMMs (Mixed Matrix Membranes), high enhancement in selectivity has been evidenced due to narrow pore aperture (3–4 æ) of MOFs. In the present case, the COFs (TpBD) contribute to a significant increase in permeability of PBI-BuI, with small deviations in selectivity. It can thus be further expected that if the COFs with even higher pore aperture are chosen, elevation in permeability could be more pronounced. In conclusion, we have showcased, for the first time, a methodology to fabricate a new class of self-supported COF@polymer hybrid membranes. These self-supported membranes are highly flexible, reproducible and display a high degree of thermal and chemical stabilities; they show potential for a wider range of applications in gas/liquid separation. As proof-of-concept, we performed gas-permeation analysis, which revealed an almost sevenfold elevation in permeability of gases (H2, N2, CH4 and CO2) compared with the pristine membrane with 50 %
4698
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication TpBD loading and appreciable CO2/N2 (23) and CO2/CH4 (48.7) selectivity. Interestingly, the effect of isoreticulation in COF towards elevating the gas permeability was prominently seen in these COF@PBI-BuI-based hybrid membranes. Overall, the findings demonstrate a versatile approach of preparing chemically stable functional COF-based hybrid membranes that could significantly contribute to the advancement of such materials in the near future. [4]
Experimental Section
[5]
TpPa-1 and TpBD were synthesized using our standard reported solvothermal syntheses method.[9] A solution casting method using DMAc as a solvent was employed for the fabrication of COF(n)@PBI-BuI hybrid membranes by varying COF content (where n = 20, 40 and 50 % by weight of TpPa-1 and TpBD). The detailed materials and experimental methods are described in the Supporting Information.
[6]
[7]
Acknowledgements
[8]
B.P.B. and H.D.C. acknowledge the UGC and CSIR (New Delhi, India) for SRF. R.B. and U.K.K. acknowledge the CSIR (CSC0122 and CSC0102), DST Indo-Singapore Project (INT/SIN/P-05) and DST Nano-mission Project (SR/NM/NS-1179/2012G) for funding.
[9]
[10]
Keywords: covalent organic frameworks · gas separation · membranes · microporous materials · pore modulation
[11]
[1] a) S. Keskin, T. M. van Heest, D. S. Sholl, ChemSusChem 2010, 3, 879; b) S. Basu, A. Cano-Odena, I. F. J. Vankelecom, J. Membr. Sci. 2010, 362, 478; c) H. B. T. Jeazet, C. Staudt, C. Janiak, Chem. Commun. 2012, 48, 2140; d) D. Nagaraju, D. G. Bhagat, R. Banerjee, U. K. Kharul, J. Mater. Chem. A 2013, 1, 8828; e) A. J. Brown, N. A. Brunelli, K. Eum, F. Rashidi, J. R. Johnson, W. J. Koros, C. W. Jones, S. Nair, Science 2014, 345, 72. [2] a) D. L. Gin, R. D. Noble, Science 2011, 332, 674; b) N. Du, H. B. Park, G. P. Robertson, M. M. Dal-Cin, T. Visser, L. Scoles, M. D. Guiver, Nat. Mater. 2011, 10, 372; c) L. M. Robeson, J. Membr. Sci. 2008, 320, 390. [3] a) B. A. Al-Maythalony, O. Shekhah, R. Swaidan, Y. Belmabkhout, I. Pinnau, M. Eddaoudi, J. Am. Chem. Soc. 2015, 137, 1754; b) S. Kim, L.
Chem. Eur. J. 2016, 22, 4695 – 4699
www.chemeurj.org
[12]
Chen, J. K. Johnson, E. Marand, J. Membr. Sci. 2007, 294, 147; c) D. Q. Vu, W. J. Koros, S. J. Miller, J. Membr. Sci. 2003, 211, 311; d) C. H. Lau, P. T. Nguyen, M. R. Hill, A. W. Thornton, K. Konstas, C. M. Doherty, R. J. Mulder, L. Bourgeois, A. C. Y. Liu, D. J. Sprouster, J. P. Sullivan, T. J. Bastow, A. J. Hill, D. L. Gin, R. D. Noble; Angew. Chem. Int. Ed. 2014, 53, 5322; Angew. Chem. 2014, 126, 5426; Angew. Chem. Int. Ed. 2014, 53, 5322; Angew. Chem. 2014, 126, 5426; e) A. F. Bushell, P. M. Budd, M. P. Attfield, J. T. A. Jones, T. Hasell, A. I. Cooper, P. Bernardo, F. Bazzarelli, G. Clarizia, J. C. Jansen, Angew. Chem. Int. Ed. 2013, 52, 1253; Angew. Chem. 2013, 125, 1291. B. Zornoza, C. Tellez, J. Coronas, J. Gascon, F. Kapteijn, Microporous Mesoporous Mater. 2013, 166, 67. a) A. P. Cút¦, A. I. Benin, N. W. Ockwig, A. J. Matzger, M. O’Keeffe, O. M. Yaghi, Science 2005, 310, 1166; b) X. Feng, X. Dinga, D. Jiang, Chem. Soc. Rev. 2012, 41, 6010. a) J. F. Dienstmaier, D. D. Medina, M. Dogru, P. Knochel, T. Bein, W. M. Heckl, M. Lackinger, ACS Nano 2012, 6, 7234; b) S. Y. Ding, J. Gao, Q. Wang, Y. Zhang, W. G. Song, C. Y. Su, W. Wang, J. Am. Chem. Soc. 2011, 133, 19816; c) S. Wan, J. Guo, J. Kim, H. Ihee, D. Jiang, Angew. Chem. Int. Ed. 2009, 48, 5439; Angew. Chem. 2009, 121, 5547. a) D. D. Hao, J. N. Zhang, H. Lu, W. G. Leng, R. L. Ge, X. N. Dai, Y. Gao, Chem. Commun. 2014, 50, 1462; b) H. Lu, C. Wang, J. Chen, R. Ge, W. Leng, B. Dong, J. Huang, Y. Gao, Chem. Commun. 2015, 51, 15562. a) L. M. Lanni, R. W. Tilford, M. Bharathy, J. J. Lavigne, J. Am. Chem. Soc. 2011, 133, 13975; b) E. Spitler, M. Giovino, S. White, W. Dichtel, Chem. Sci. 2011, 2, 1588. a) S. Kandambeth, A. Mallick, B. Lukose, V. M. Mane, T. Heine, R. Banerjee, J. Am. Chem. Soc. 2012, 134, 19524; b) S. Chandra, S. Kandambeth, B. P. Biswal, B. Lukose, S. M. Kunjir, M. Chaudhary, R. Babarao, T. Heine, R. Banerjee, J. Am. Chem. Soc. 2013, 135, 17853. a) T. Yang, Y. Xiao, T. S. Chung, Energy Environ. Sci. 2011, 4, 4171; b) S. C. Kumbharkar, K. Li, J. Membr. Sci. 2012, 415, 793; c) A. Bhaskar, R. Banerjee, U. K. Kharul, J. Mater. Chem. A 2014, 2, 12962; d) S. C. Kumbharkar, P. B. Karadkar, U. K. Kharul, J. Membr. Sci. 2006, 286, 161. a) G. R. Desiraju, Acc. Chem. Res. 2002, 35, 565; b) G. R. Desiraju, T. Steiner, The Weak Hydrogen Bond in Structural Chemistry and Biology, Oxford University Press, Oxford, 1999. a) W. F. Yong, F. Y. Li, Y. C. Xiao, P. Li, K. P. Pramod, Y. W. Tong, T. S. Chung, J. Membr. Sci. 2012, 47, 407; b) J. S. McHattie, W. J. Koros, D. R. Paul, Polymer 1992, 33, 1701; c) M. W. Hellums, W. J. Koros, G. R. Husk, D. R. Paul, J. Membr. Sci. 1989, 46, 93.
Received: December 2, 2015 Published online on February 18, 2016
4699
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
