Synthesis of 2D Imine-Linked Covalent Organic Frameworks through

Aug 30, 2017 - Covalent organic frameworks (COFs) are crystalline, permanently porous, two-dimensional or three-dimensional polymers with tunable topo...
48 downloads 23 Views 2MB Size
Communication pubs.acs.org/JACS

Synthesis of 2D Imine-Linked Covalent Organic Frameworks through Formal Transimination Reactions Edon Vitaku and William R. Dichtel* Department of Chemistry, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States S Supporting Information *

ABSTRACT: Covalent organic frameworks (COFs) are crystalline, permanently porous, two-dimensional or threedimensional polymers with tunable topology and functionality. COFs linked with imines or β-ketoenamines are more chemically stable than their boron-linked counterparts, making them more promising for a broad range of applications, including energy storage devices, proton-conductive membranes, and catalyst supports. We report a general and scalable method for synthesizing imine- and β-ketoenamine-linked COFs based on the formal transimination of N-aryl benzophenone imines. These substrates are often the synthetic precursors of traditional polyfunctional aryl amine monomers and are more stable, soluble, and easy to handle and purify. The imine- and β-ketoenamine-linked COFs obtained from this approach show excellent materials quality, as characterized by X-ray diffraction and surface area analysis. The most optimized COF exhibited a Brunauer−Emmett−Teller surface area (>2600 m2/g) very close to its theoretical value (2830 m2/g). This method is amenable to both conventional solvothermal conditions and microwave heating, providing similar or even improved materials quality with shorter reaction times. The high materials quality, scalability, and availability of benzophenone imine monomers are all attractive features of this approach.

Figure 1. Comparison of the synthesis of imine-linked 2D COFs from polyfunctional aryl amine monomers (left) and the corresponding benzophenone imines (right).

none imine.7 The resulting N-aryl benzophenone imines are then hydrolyzed to the corresponding free amines. Recognizing that imine-linked COFs crystallize under conditions that promote dynamic imine exchange,8 we hypothesized that Naryl benzophenone imines are competent monomers for COF formation (Figure 1). Compared to the corresponding free amines, N-aryl benzophenone imines are typically more soluble, less prone to oxidation, and more easily purified. Their use in COF synthesis also precludes the need for a separate deprotection step. We prepared both imine-linked and βketoenamine-linked two-dimensional (2D) COFs from benzophenone-imine protected monomers. These 2D COFs exhibit excellent crystallinity and Brunauer−Emmett−Teller (BET) surface areas higher than networks derived from the corresponding aryl amines. Our method is compatible with both traditional solvothermal conditions and microwave irradiation, which require shorter reaction times. These results demonstrate that N-aryl benzophenone imines are desirable

C

ovalent organic frameworks (COFs)1 are porous, crystalline networks made from light elements such as H, B, C, N, and O. COFs offer well-defined solid-state structures and tunable pore sizes that emerge from the topology and shape of their molecular building blocks. The first reported COFs were linked by boronate esters and boroxines, but these linkages are easily hydrolyzed or oxidized. Therefore, iminelinked and other nitrogen-containing COFs show greater promise for emerging applications,2 such as energy storage,3 catalysis,4 and drug delivery.5 Imine-linked COFs are typically prepared through the condensation of aryl amines and aldehydes under Brønsted acid-catalyzed solvothermal conditions. We recently reported that metal triflate catalysts also provide imine-linked COFs with excellent crystallinity and high surface areas under exceptionally mild conditions.6 The limited availability and poor oxidative stability of many polyfunctional aryl amines are major stumbling blocks for the design and synthesis of new imine-linked COFs. Monomers that are not commercially available are commonly prepared via a Buchwald−Hartwig coupling of aryl halides with benzophe© 2017 American Chemical Society

Received: July 3, 2017 Published: August 30, 2017 12911

DOI: 10.1021/jacs.7b06913 J. Am. Chem. Soc. 2017, 139, 12911−12914

Communication

Journal of the American Chemical Society

The resulting BND-TFB COF was activated by washing with supercritical CO2, giving an SBET of 1938 m2/g (entry 1, see SI for a pictorial guide).10 Extending the reaction time to 5 days (entries 2−4) increased the SBET to 2618 m2/g, which is among the highest reported values of a 2D COF. The powder X-ray diffraction pattern of the isolated material is consistent with an approximately eclipsed model structure (Figure 2A−C), which exhibits a calculated Connolly surface area of 2830 m2/g (Figure 2E,F). Scherrer analysis revealed a domain size of approximately 25 nm (see Table S5 and Figure S66). Analysis of the N2 adsorption isotherm using nonlocal density functional theory is consistent with a narrow pore size distribution in good agreement with the calculated pore size of 24 Å (Figure 2D).11 BND-TFB COF was also characterized by infrared spectroscopy, elemental analysis, 13C cross-polarization magic angle spinning NMR spectroscopy, and scanning electron microscopy (see Supporting Information), which were consistent with its expected structure and suggested its phase purity. We evaluated modifications to the newly established conditions (i.e., entries 1−4) based on their impact on the SBET of the isolated BND-TFB COF, as surface area is a quantitative parameter that is highly sensitive to the polymerization and activation conditions. These experiments demonstrated that the BND-TFB COF may be activated under dynamic vacuum instead of supercritical CO2 (entry 5), and that thorough deoxygenation of the polymerization mixture is unnecessary (entry 6). When the reaction scale was increased 15-fold (entry 7), a COF with no difference in crystallinity or surface area was isolated. However, polymerizations conducted under reduced pressure, which is common practice, yielded COFs with decreased surface area (entry 8). These results suggest that the monomers and reaction intermediates do not undergo noticeable oxidative degradation, which enables the use of simple reaction setups that do not require degassing procedures. We also extended this method to microwave irradiation,12 which allows for COF formation of similar materials quality in shorter times (entries 9−12). A 1 h reaction time yielded a material with a surface area of only 629 m2/g

Table 1. Evaluation of BND-TFB COF Synthetic Procedures

entry

deviation from above conditions

time

SBET

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

− − − − vacuum activation under air 15× scale, vacuum activation under vacuum microwave, under air microwave, under air microwave, under air microwave, under air 3 used (instead of 2) 3 used with 2 equiv 4 3 used, under air

24 h 2d 3d 5d 3d 3d 3d 3d 1h 3h 5h 7h 3d 3d 3d

1938 m2/g 2396 m2/g 2314 m2/g 2618 m2/g 2313 m2/g 2327 m2/g 2321 m2/g 1677 m2/g 629 m2/g 1558 m2/g 2100 m2/g 2200 m2/g 1501 m2/g 1453 m2/g 1381 m2/g

monomers, particularly when incorporating synthetically advanced building blocks into these periodic polymer networks. 1,3,5-Triformylbenzene (1) and N-benzidine benzophenone imine (2) were condensed under solvothermal conditions to obtain an imine-linked network, BND-TFB COF (Table 1).2a This COF was selected because it has been prepared previously from benzidine (3), which is prone to oxidation, with a maximum BET surface area (SBET) of 1496 m2/g.5b,9 A deoxygenated mixture of 1 and 2 in a mixture of mesitylene, 1,4-dioxane, and 6 M aqueous CH3CO2H (3:3:1) was sealed under a nitrogen atmosphere and heated to 120 °C for 24 h.

Figure 2. BND-TFB COF. (A) Top view, (B) side view. (C) Experimental, Pawley-refined, predicted powder X-ray diffraction patterns, and a difference plot (experimental pattern minus refined pattern), (D) pore size distribution, and (E) nitrogen isotherms (C−E correspond to Table 1, entry 3). (F) Summary of the obtained BET surface areas through microwave heating (Table 1, entries 9−12), conventional heating (Table 1, entries 1−4), and its calculated Connolly surface area. 12912

DOI: 10.1021/jacs.7b06913 J. Am. Chem. Soc. 2017, 139, 12911−12914

Communication

Journal of the American Chemical Society (entry 9). This value increased to 2200 m2/g (entry 12) after 7 h, which indicated an optimal irradiation time between 5 and 7 h. To determine whether the role of benzophenone is to modulate the number of available amine functionalities through slow deprotection13 or act as a competitor,14 we performed two control experiments. A direct condensation between free benzidene (3) and 1 was performed in the absence (entry 13) or presence (entry 14) of 2 equiv of benzophenone (4). Both afforded a COF with a lower SBET than COFs derived from 2 (entry 3). This result suggests that free benzophenone does not act as a competitor. The presence of air also has a minor effect on the SBET (entry 15). Thus, we hypothesize that the slow deprotection of the benzophenone imine protecting groups, which modulates the available free amine functionalities, is responsible for the enhanced COF quality. With the optimized conditions for both conventional heating (entries 3 and 4) and microwave irradiation (entry 11) established, we evaluated the synthesis of a second iminelinked linked COF (Table 2). The TAPB-PDA COF (entry 2), which has a pore size larger than that of BND-TFB COF, was also isolated as crystalline network (Figure 3A) with a higher BET surface area (Figure 3C) than previously reported under solvothermal conditions.8,15 The pore size distributions for both samples agreed with the theoretical values (Figure 3B, top). In addition, Fourier transform infrared (Figure S70) spectra showed no evidence of benzophenone contamination in any of the isolated COFs. We also evaluated the application of this method to βketoenamine-linked COFs, which condense via a different mechanism than for imine-linked COFs.2 Imine-linked COFs undergo continuous error correction because of the reversible

Figure 3. Characterization data for COFs obtained under conventional heating. (A) PXRD data, (B) pore size distributions, and (C) nitrogen isotherms.

nature of the imine bond in the presence of aqueous acid.8 In contrast, β-ketoenamine-linked COFs undergo two potentially reversible imine-bond-forming steps followed by tautomerization to the β-ketoenamine structure after the third imine is formed.16 Given the notable stability of the final COF materials to aqueous acid, it is unlikely that these networks correct defects through hydrolysis after this tautomerization has occurred, and β-ketoenamine-linked COFs have not yet exhibited as high crystallinity or SBET as their imine-linked counterparts. Nevertheless, their outstanding stability makes these COFs of great applied interest.3a,c,17 Both conventional heating and microwave irradiation yielded DAB-TFP and BNDTFP COFs with BET surface areas higher than those of all previous reports (entries 3 and 4; see Figure 3C).12b,18 The powder X-ray diffraction patterns and pore size distributions of these materials were also in good agreement with modeled structures (see Figure 3A,B). The effectiveness of the benzophenone imine-based procedure for these materials is consistent with our previous optimized synthesis of a redoxactive β-ketoenamine-linked COF, in which crystalline thin films were only obtained upon slow addition of TFP to a solution of the diamine.13 It is possible that the continuous deprotection of the benzophenone imine-protected monomers achieves improved materials quality through a similar effect. In conclusion, we developed a practical, general, and scalable method based on benzophenone imine-containing monomers for preparing nitrogen-linked COFs with BET surface areas superior to those of traditional syntheses. This approach is amenable to conventional solvothermal conditions or microwave irradiation and allows access to high-quality COFs within reaction times as short as 5 h. The N-aryl benzophenone imine monomers are typically more soluble and oxidatively stable

Table 2. COF Scope Evaluation under Conventional Heating and Microwave Irradiation

12913

DOI: 10.1021/jacs.7b06913 J. Am. Chem. Soc. 2017, 139, 12911−12914

Communication

Journal of the American Chemical Society

(4) (a) Ding, S. Y.; Gao, J.; Wang, Q.; Zhang, Y.; Song, W. G.; Su, C. Y.; Wang, W. J. Am. Chem. Soc. 2011, 133, 19816−19822. (b) Xu, H.; Chen, X.; Gao, J.; Lin, J. B.; Addicoat, M.; Irle, S.; Jiang, D. L. Chem. Commun. 2014, 50, 1292−1294. (c) Lin, S.; Diercks, C. S.; Zhang, Y. B.; Kornienko, N.; Nichols, E. M.; Zhao, Y. B.; Paris, A. R.; Kim, D.; Yang, P.; Yaghi, O. M.; Chang, C. J. Science 2015, 349, 1208−1213. (d) Xu, H.; Gao, J.; Jiang, D. L. Nat. Chem. 2015, 7, 905−912. (5) (a) Vyas, V. S.; Vishwakarma, M.; Moudrakovski, I.; Haase, F.; Savasci, G.; Ochsenfeld, C.; Spatz, J. P.; Lotsch, B. V. Adv. Mater. 2016, 28, 8749−8754. (b) Bai, L. Y.; Phua, S. Z. F.; Lim, W. Q.; Jana, A.; Luo, Z.; Tham, H. P.; Zhao, L. Z.; Gao, Q.; Zhao, Y. L. Chem. Commun. 2016, 52, 4128−4131. (c) Mitra, S.; Sasmal, H. S.; Kundu, T.; Kandambeth, S.; Illath, K.; Diaz, D. D.; Banerjee, R. J. Am. Chem. Soc. 2017, 139, 4513−4520. (6) Matsumoto, M.; Dasari, R. R.; Ji, W.; Feriante, C. H.; Parker, T. C.; Marder, S. R.; Dichtel, W. R. J. Am. Chem. Soc. 2017, 139, 4999− 5002. (7) Wolfe, J. P.; Ahman, J.; Sadighi, J. P.; Singer, R. A.; Buchwald, S. L. Tetrahedron Lett. 1997, 38, 6367−6370. (8) Smith, B. J.; Overholts, A. C.; Hwang, N.; Dichtel, W. R. Chem. Commun. 2016, 52, 3690−3693. (9) (a) Gao, Q.; Bai, L. Y.; Zeng, Y. F.; Wang, P.; Zhang, X. J.; Zou, R. Q.; Zhao, Y. L. Chem. - Eur. J. 2015, 21, 16818−16822. (b) Bai, L. Y.; Gao, Q.; Zhao, Y. L. J. Mater. Chem. A 2016, 4, 14106−14110. (10) The plots chosen for the BET calculcation were distributed in the range of 0.05−0.1 P/P0. Although SBET values can vary with the relative pressure range chosen for calculation, we consistently found that the highest correlation coefficients were obtained when the BET plots were distributed in that range. (11) The pore size was previously reported to be 19Å (see ref 9). Our Pawley refined structure yielded a pore size of 24 Å. (12) (a) Campbell, N. L.; Clowes, R.; Ritchie, L. K.; Cooper, A. I. Chem. Mater. 2009, 21, 204−206. (b) Wei, H.; Chai, S. Z.; Hu, N. T.; Yang, Z.; Wei, L. M.; Wang, L. Chem. Commun. 2015, 51, 12178− 12181. (13) DeBlase, C. R.; Hernandez-Burgos, K.; Silberstein, K. E.; Rodriguez-Calero, G. G.; Bisbey, R. P.; Abruna, H. D.; Dichtel, W. R. ACS Nano 2015, 9, 3178−3183. (14) (a) Calik, M.; Sick, T.; Dogru, M.; Doblinger, M.; Datz, S.; Budde, H.; Hartschuh, A.; Auras, F.; Bein, T. J. Am. Chem. Soc. 2016, 138, 1234−1239. (b) Smith, B. J.; Hwang, N.; Chavez, A. D.; Novotney, J. L.; Dichtel, W. R. Chem. Commun. 2015, 51, 7532−7535. (15) Waller, P. J.; Lyle, S. J.; Popp, T. M. O.; Diercks, C. S.; Reimer, J. A.; Yaghi, O. M. J. Am. Chem. Soc. 2016, 138, 15519−15522. (16) Kandambeth, S.; Mallick, A.; Lukose, B.; Mane, M. V.; Heine, T.; Banerjee, R. J. Am. Chem. Soc. 2012, 134, 19524−19527. (17) (a) Chandra, S.; Kundu, T.; Kandambeth, S.; BabaRao, R.; Marathe, Y.; Kunjir, S. M.; Banerjee, R. J. Am. Chem. Soc. 2014, 136, 6570−6573. (b) Wang, S.; Wang, Q. Y.; Shao, P. P.; Han, Y. Z.; Gao, X.; Ma, L.; Yuan, S.; Ma, X. J.; Zhou, J. W.; Feng, X.; Wang, B. J. Am. Chem. Soc. 2017, 139, 4258−4261. (18) (a) Biswal, B. P.; Chandra, S.; Kandambeth, S.; Lukose, B.; Heine, T.; Banerjee, R. J. Am. Chem. Soc. 2013, 135, 5328−5331. (b) Pachfule, P.; Kandambeth, S.; Diaz, D. D.; Banerjee, R. Chem. Commun. 2014, 50, 3169−3172. (c) Biswal, B. P.; Kandambeth, S.; Chandra, S.; Shinde, D. B.; Bera, S.; Karak, S.; Garai, B.; Kharul, U. K.; Banerjee, R. J. Mater. Chem. A 2015, 3, 23664−23669. (d) Tan, J.; Namuangruk, S.; Kong, W. F.; Kungwan, N.; Guo, J.; Wang, C. C. Angew. Chem., Int. Ed. 2016, 55, 13979−13984. (19) Alahakoon, S. B.; McCandless, G. T.; Karunathilake, A. A. K.; Thompson, C. M.; Smaldone, R. A. Chem. - Eur. J. 2017, 23, 4255− 4259.

than the corresponding free amines. This method was used to obtain a BND-TFB COF with a BET surface area that closely approaches its theoretical value, narrow pore size distribution, and high crystallinity. Although improved materials quality in COFs has been achieved by incorporating fluorines19 or methoxy groups4d ortho to aldehyde moieties, our results demonstrate that these design criteria may be beneficial, but not required, to obtain high-quality COFs.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b06913. Experimental procedures, additional PXRD patterns, nitrogen isotherms, BET plots, pore size distributions, FTIR spectra, elemental analyses, SEMs, NMR spectra, and pictorial guide to covalent organic framework synthesis (PDF) Crystallographic data (CIF)



AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Edon Vitaku: 0000-0002-0057-4245 William R. Dichtel: 0000-0002-3635-6119 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We acknowledge the Army Research Office for a Multidisciplinary University Research Initiatives (MURI) award under grant number W911NF-15-1-0447. W.R.D. was also supported by the Camille and Henry Dreyfus Foundation through a Camille Dreyfus Teacher-Scholar Award. E.V. acknowledges Ryan Bisbey and Anton Chavez for helpful discussions. This study made use of the IMSERC and EPIC at Northwestern University, both of which have received support from the Soft and Hybrid Nanotechnology Experimental (SHyNE) Resource (NSF NNCI-1542205 and NSF ECCS1542205, respectively), the State of Illinois, and the International Institute for Nanotechnology (IIN).



REFERENCES

(1) For selected reviews and perspectives on COFs, see: (a) Feng, X.; Ding, X. S.; Jiang, D. L. Chem. Soc. Rev. 2012, 41, 6010−6022. (b) Ding, S. Y.; Wang, W. Chem. Soc. Rev. 2013, 42, 548−568. (c) Colson, J. W.; Dichtel, W. R. Nat. Chem. 2013, 5, 453−465. (d) Xiang, Z. H.; Cao, D. P.; Dai, L. M. Polym. Chem. 2015, 6, 1896− 1911. (e) Waller, P. J.; Gandara, F.; Yaghi, O. M. Acc. Chem. Res. 2015, 48, 3053−3063. (f) Huang, N.; Wang, P.; Jiang, D. L. Nat. Rev. Mater. 2016, 1, 16068. (g) Diercks, C. S.; Yaghi, O. M. Science 2017, 355, eaal1585. (h) Bisbey, R. P.; Dichtel, W. R. ACS Cent. Sci. 2017, 3, 533−543. (2) (a) Segura, J. L.; Mancheno, M. J.; Zamora, F. Chem. Soc. Rev. 2016, 45, 5635−5671. (b) DeBlase, C. R.; Dichtel, W. R. Macromolecules 2016, 49, 5297−5305. (3) (a) DeBlase, C. R.; Silberstein, K. E.; Truong, T. T.; Abruna, H. D.; Dichtel, W. R. J. Am. Chem. Soc. 2013, 135, 16821−16824. (b) Xu, F.; Jin, S.; Zhong, H.; Wu, D.; Yang, X.; Chen, X.; Wei, H.; Fu, R.; Jiang, D. Sci. Rep. 2015, 5, 8225. (c) Mulzer, C. R.; Shen, L. X.; Bisbey, R. P.; McKone, J. R.; Zhang, N.; Abruna, H. D.; Dichtel, W. R. ACS Cent. Sci. 2016, 2, 667−673. 12914

DOI: 10.1021/jacs.7b06913 J. Am. Chem. Soc. 2017, 139, 12911−12914