Covalent Organic Frameworks ( COF S ) : A Review

The search for supramolecular promising porous crystalline materials with diverse applications such as gas storage, catalysis, chemo-sensing, energy storage, and optoelectronic have led to the design and construction of Covalent Organic Frameworks (COFs). COFs are a class of porous crystalline polymers that allow the precise integration of organic building blocks and linkage motifs to create predesigned skeletons and nano-porous materials. In this review article, a historic overview of the chemistry of COFs, survey of the advances in topology design and synthetic reactions, basic design principles that govern the formation of COFs as porous crystalline polymers as well as common synthetic procedures and characterization techniques are discussed. Furthermore some challenges associate with the synthesis of COFs are highlighted. We hope that this review will help researchers, industrialists and academics in no mean feat. DOI: https://dx.doi.org/10.4314/jasem.v26i1.22 Open Access Article: (https://pkp.sfu.ca/ojs/) This an open access article distributed under the Creative Commons Attribution License (CCL), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Impact factor: http://sjifactor.com/passport.php?id=21082 Google Analytics: https://www.ajol.info/stats/bdf07303d34706088ffffbc8a92c9c1491b12470 Copyright: Copyright © 2022 Bull et al. This is an open access article (https://pkp.sfu.ca/ojs/) distributed under the Creative Commons Attribution License (CCL), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Dates: Received: 23 August 2021; Revised: 21 December 2021; Accepted: 06 January 2022 .

the linking of well-defined organic building units with diverse functionalities through strong covalent bonding to give 1-D, 2-D or 3-D structures. Research in this area of chemistry was initially not very promising due to difficulties associated with the synthesis of crystalline COFs for X-ray characterization. ) However, the shortcomings of COFs' crystallinity has been partly overcome by modifying synthetic methods and the emergence and application of reticular chemistry, which makes use of topologically designed building units for the construction of crystalline COFs via the use of boronic acids or boronate esters as precursors. (Jinang, 2012) Yaghi and co-workers used topological design principles to successfully prepare the first examples of COFs in 2005. (Ockwig et al., 2005a) With this breakthrough, there has been rapid progress in COF chemistry in many parts of the globe as COFs display great potential for functional applications. COFs are constructed by using well defined organic building blocks, containing bonds such as C-C, C-O, C-N, C=N, C-Si, B-O and B-N. (Ockwig et al., 2005a;Saha et al., 2013;Wan et al., 2008) The reactions leading to the formation of COFs are usually reversible condensation reactions, such as the formation of B-O bonds as seen in boronate, boroxines and borosilicate, Hunt et al., 2008) C-N bonds formed in triazine and imidazation, (Chem et al., 2012;Sha Lin et al., 2015; B-N bonds formed in boraxines, (Jackson et al., 2012;Spitler & Dichtel, 2010) C=N bonds formed in imine, hydrazine and squaraine, Uribe-romo et al., 2009Uribe-romo et al., , 2011Zhou et al., 2014) as well as N-N bonds formed in azodioxides. (Beaudoin et al., 2013) In comparison to other porous materials ( zeolites and MOFs), COFs have the advantage of low density, high thermal and chemical stabilities, permanent porosity, tunable pore size and structure, columnar π-stacking structures as well as versatilities in covalent combination of building units. Gao et al., 2018;Hisaki et al., 2017;N. Huang, Zhai, et al., 2016) The columnar stacking structure in some COFs provides a unique avenue to constructing ordered π-systems that are difficult if not impossible to achieve with conventional covalent or non-covalent approaches. These unique features of COFs have attracted considerable interests amongst scientists, engineers and industrialists thus making COFs novel materials for important applications such as gas adsorption, (Babarao & Jiang, 2008;Cui et al., 2017;Doonan et al., 2010;Jin-Tao Yu;Zhe Chen;Junliang Sun;Zhi-Tang Huang;Qi-Yu Zheng, 2012;Mendoza-Cortés et al., 2012;Yang et al., 2018;Bull et al. (2020)) optoelectronics and photovoltaics, (Calik et al., 2014;Dogru et al., 2013;Jiang, 2014;Xuan-he Liu et al., 2014;Wan et al., 2008Wan et al., , 2011 gas separation, (Lan et al., 2017;Ma, Heping;Ren, Hao;Meng, Shuang;Yan, Zhuojun;Zhao, Huanyu;Sun, Fuxing;Oh et al., 2013) catalysis, (Article et al., 2014;He et al., 2015;Song Lin et al., 2015;Pachfule et al., 2018) proton conductivity,(X. Guo et al., 2017;Kunjir and Banerjee, 2014;J. Li et al., 2018) chemical sensor, (Dalapati et al., 2013;Das, Gobinda;Biswal P, Bishnu;Kandambeth, Sharath;Kaur, Gagandeep;addicoat, Matthew;Heine, Thomas;Nagai et al., 2013) drug delivery, (Guo et al., 2015) energy storage (Deblase et al., 2013;Xu et al., 2015) as well as in chromatographic separations.{Formatting Citation}. In recent times, great efforts have been directed towards the enhancement of COFs for CO2 capture and rapid research and advancement has been reported. Hug et al., 2015;Olajire, 2017;Zeng et al., 2016) However, there is still more to be done concerning COFs for CO2 capture and purification as well as construction of COFs with large CO2 adsorption and selectivity over other gases in the flue gas stream. (Olajire, 2017) In addition to the progress made in COF chemistry for CO2 adsorption, the need for high thermal and chemical stabilities and the recyclability of COF materials without loss of performance in air are still major problems. Thus, new synthetic strategies, enhancement of crystallinity, stability and porosity, and CO2 capture efficiency as well as improved selectivity with respect to other flue gas are all factors of COFs that need improvement.

COF Design
Dynamic Covalent Chemistry of COFs): Kinetic control of reactions leading to the synthesis of polymeric material is key to irreversible formation of COFs (containing mainly covalent bonds). (Diercks and Yaghi, 2017) It has always been a challenge to crystallize linked organic polymers into solids by means of irreversible reactions. Uribe-romo et al., 2009) However, dynamic covalent chemistry which involves the breaking /formation bonds leads to the reversible formation of covalent bonds. Dynamic covalent bond chemistry is thermodynamically driven, as a result, it offers reversible reactions process with ''error checking'' as well as ''proof reading'' features and thus leads to the formation of the most thermodynamically stable products. Thus, the application of dynamic covalent chemistry in COFs enables the polymer skeleton formation to occur alongside the crystallization process, while self-healing feedback reduces the incidence of structural defects and helps the formation of an ordered structure.(N. Huang, Wang, et al., 2016;N. Huang, Zhai, et al., 2016;Jinang, 2012;Smith et al., 2017) The end effect of the dynamic covalent chemistry is that the final COF product has an ordered crystalline structure having high thermodynamic stability. To design and synthesise a COF, two key factors must be considered in order to achieve thermodynamic control in reversible reactions. (Diercks and Yaghi, 2017;Jinang, 2012) These factors are: the functionality and structure of the building units, and the synthetic methodology (consideration of the reaction media and the reaction conditions are important).

Structure of Building Units:
The structure of the building unit for COF construction must meet two requirements. Firstly, the reaction leading to the formation of the COF material should be a reversible reaction. This implies that the building units should be ones that undergo dynamic covalent bond formation. That is, there should be no irreversible reactions and the reaction system should contain only monomers, oligomers and polymers that are interchangeable under thermodynamic conditions. Secondly, the geometry of the secondary building units should be retained in the COF. This means that rigid building units as well as discrete directional bond formation are required. In a bid to satisfy the first requirement, the chemistry of boronic acids, triazines and imines has been explored for successful COFs syntheses. (Beuerle & Gole, 2018;Díaz & Corma, 2016;Jinang, 2012;San-Yuan Ding, 2013;Zeng et al., 2016) Building Blocks for COFs Boron-based COFs: A large number of known COFs depend on boron chemistry because boronic acids can undergo condensation reaction through either selfcondensation or co-condensation with diols to give six-membered boroxine and five-membered boronate ester rings as shown Figure 1.
Scheme 2: Synthetic scheme and proposed structure for COF-18Å. (Tilford et al., 2006) Scheme 3: Synthesis of COF-66. (Wan et al., 2011) From the Table 1 above, it can be seen that the SBET of the 2-D COFs decreases in the order: COF-10 > COF-5 COF-8 > >COF-6 = COF-1> COF-66. But the CO2 uptake capacities of these COFs at 273 K/1 bar decreases as follows: COF-6 > COF-1 > COF-8 > COF-5 > COF-10. Thus irrespective of the fact that COF-6 has a pore size of 0.9 nm as well a SBET of 750 m 2 /g, it has the highest CO2 uptake capacity of 167 mg/g at 273 K/ 1bar compared to the other 2-D COF in the Table 1. This suggests that the CO2 uptake of the COFs at low pressure depends more on the pore diameter than on the SBET. However, at a higher pressure of 55 bar and 278 K, the order of the CO2 uptake capacities of these COFs is: COF-10 (1010 mg/g) > COF-5 (870 mg/g) >COF-8 (630 mg/g) > COF-6 (310 mg/g) > COF-1 (230 mg/g). This implies that at high pressures the CO2 uptake capacities of these COFs is more related to their SBET than on their pore sizes. In addition to the list of 2-D boron-based COFs, Kahvechi et al. successfully synthesised a 2-D mesoporous triptycene-derived COF (TDCOF-5) using 1,4-benzenediboronic acid and hexahydroxytriptycene solvothermally to give a boronate as shown in Scheme 2.3.  The SBET and pore volume reported for TDCOF-5 were 2497 m 2 /g and 1.3 cm 3 /g at p/po of 0.9. The SBET for TDCOF-5 is higher than that of COF-5 (1670 m 2 /g). (Díaz & Corma, 2016;Olajire, 2017) The reported CO2 adsorption capacity of TDCOF-5 was 92 mg/g at 273 K/ 1 bar.   Scheme 4: Synthesis OF COF TDCOF-5.  The rigid conformation of the building units drives the topological design of the COFs. The rigid nature as well as discrete bonding direction of arene molecules makes aromatic π-systems good building units. The diversity of π-aromatic systems allows the combination of various building units which, results in the formation of COFs with very variable molecular design. (Jinang, 2012) Some building blocks used for the synthesis of boron-based COFs are summarised below Figure 2. These building blocks can be classified as either 2-D or 3-D on the basis of the directional symmetry of the coordinating groups. The geometry of the building units determines the final structure of the COF. There are fewer 3-D COFs than 2-D COFs due to the limited number of building blocks available. Thus the self-condensation of tetrahedral nodes (3D-T4) or co-condensation of tetrahedral nodes with linear 2D-C3 or triangular 2D-C3 building units can lead to the construction of 3-D COFs with different crystalline space groups. However, a combination of 2-D (2D-C2 + 2D-C3, 2D-C3 + 2D-C3, or 2D-C2 + 2D-C4) will lead to formation of 2-D COFs. (Jinang, 2012     The SBET, densities and CO2 uptake capacities at low pressure of these 3-D COFs are summarised in Table 2. Another boronbased 3-D COF is COF-108 with a bor topology and a pore size of 3.1 nm and low density of 0.17 g/cm 3 . Jinang, 2012;Olajire, 2017) COF-108 was synthesised via the co-condensation of tetrakis(4-dihydroxyborylphenyl)silane (TDBPS) and triphenylene-2,3,6,7,10,11-hexaol (TPH) as shown below in Scheme 6. .

Triazine-based COFs (CTF)
The pioneers of the use of cyano-aromatic compounds for the synthesis of triazine-based crystalline COF (COF-CTF-1) were Thomas and co-workers (2008).  The synthesis of CTF-1 was done by the reversible cyclotrimerization reaction of aromatic nitrile building unit (1,4-dicyanobenzene) in the presence of ZnCl2 which served as a catalyst as well as a solvent at 400 °C as shown in Scheme 7  Scheme 7: Synthesis of 2D triazine-based COF (CTF-1).  Scheme 8: Trimerization of 1,3,5-tricyanobenzene to CTF-0. (Katekomol et al., 2013) CTF-1 has a 2-D hexagonal COF structure with pore size of 1.2 nm, 1-D pore channels and SBET of ca. 1000 m 2 /g. In addition, Thomas and co-workers (Katekomol et al., 2013) also synthesised CTF-0 via the trimerization of 1,3,5-tricyanobenzene in molten ZnCl2 as shown in Scheme 8. It was reported that the crystalline framework (CTF-0) possessed a SBET of 500 m 2 /g and CO2 uptake capacity of 186 mg/g at 273 K, while CTF-0 derivatives obtained at higher temperatures had an increased SBET of 2000 m 2 /g.
The transformation of tetrakis(4cyanophenyl)ethylene under Lewis acidic (ZnCl2) conditions at 400 °C was studied for different ratios of tetrakis(4-cyanophenyl)ethylene and ZnCl2 to give two porous covalent triazine-based frameworks (PCTF-1 & PCTF-2, which have the same overall structure but a different porosity dependent on the reaction conditions) with SBET of 2235 and 784 m 2 /g respectively (Bhunia, Vasylyeva, et al., 2013) Scheme 9.
Scheme 11: Structures of PCTF-1, PCTF-2, PCTF-3 & PCTF-4.  The SBET of these four triazine-based COFs was reported to decrease as the length of the branched arm elongates. Thus PCTF-1 Scheme 11 ( containing biphenyl as the branched arm, PCTF-2 with terphenyl branched arm and PCTF-3 with four-phenyl branched arm have SBET 853, 811 and 395 m 2 /g, respectively. It was reported that PCTFs containing monomers with longer branches packed more efficiently, which resulted in higher density and lower surface areas. However, comparing PCTF-4 containing benzothiadiazole in the place of the central benzene in PCTF-2, the N2 adsorption isotherms showed its SBET to be 1404 m 2 /g, which ranked highest among the PCTFs, almost twice that of PCTF-2. The three nitrogen atoms in the C3N3 triazine rings in them, which are reported to be responsible for the high affinity of these frameworks toward the adsorption of CO2. The CO2 uptake capacity reported for PCTF-4 was 205 mg/g at 273 K/ 1 bar, one of the highest reported for this class of CTF materials. The high CO2 adsorption capacity of PCTF-4 is attributed to the presence of the benzothiadiazole in the framework. The benzothiadiazole served as a polar group that enhanced the affinity of the framework toward CO2. This observation suggests that the incorporation of strong polar groups such as benzothiadiazole into the framework of a microporous polymer could improve the binding affinity of CO2 molecules towards the polymer. PTCF-4, was also reported to have CO2/N2 and CO2/CH4 selectivity ratios ranging from 14-56 and 11-20 respectively, at 273 K. . Hydrocarbons containing fluorine atoms are more hydrophobic and lipophilic than corresponding hydrocarbons lacking fluorine and show stronger affinity toward CO2 than hydrocarbons without fluorine substituents. Similar observations of the positive effects concerning fluorine atoms in framework materials were reported by Smaldone and co-workers (Alahakoon et al., 2017) who reported that increasing the ratio of non-fluorinated to fluorinated monomers led to significant improvement in both crystallinity and porosity of TFx-COFs (Alahakoon et al., 2017). With this in mind, Han and co-workers (Y.  designed and constructed two triazine-based covalent frameworks (CTF-1 & FCTF-1) as shown in Scheme 12 . The target for synthesis of FCTF-1 was to prepare a material with high water tolerance and a strong preferences for binding CO2. (Y.  Scheme 12: Synthesis of CTF-1 and perfluorinated covalent triazine-based framework (FCTF-1)(Y.  BULL, OS; BULL, I; AMADI, GK; ODU, CO The addition of fluorine in FCTF-1 was reported to create an F-rich framework which enhanced the CO2 adsorption capacity of FCTF-1 in comparison to CTF-1 without fluorine atoms. The fluorine atoms in the framework are highly electronegative and thus produce strong polar covalent C-F bonds which increase CO2 electrostatic interactions. In addition, it was reported that the presence of F atoms in the PCTF-1 compared to CTF-1 reduced the pore size considerably to <0.5 nm, thus increasing the high CO2 uptake capacity as well as the kinetic selectivity for CO2/N2 separation. In the mixed-gas breakthrough experiments, FCTF-1 was reported to display a remarkable CO2/N2 selectivity ratio of 77 under kinetic flow conditions. The ratio 77 is much higher than predicted selectivity ratio of 31 from single-gas equilibrium uptake data. Most framework materials tend to perform poorly in their adsorption of CO2 in the presence of moisture. However, the performance of FCTF-1 in terms of CO2 adsorption was also reported to remain excellent even in the presence of moisture in the flue gas stream. This was attributed to the hydrophobic nature of the C-F groups. (Olajire, 2017;Y. Zhao et al., 2013). Although, both CTF-1 and FCTF-1 were reported to have poor crystallinity, they were heated at 600 °C to give CTF-1-600 and PCTF-1-600 respectively. However, no fluorine atoms were detected in FCTF-1-600, which suggested that F species had been removed from the polymer framework by thermal decomposition at high temperatures. At 600 °C both FCTF-1 and FCTF-1-600 were reported to display type-I isotherms. The CO2 adsorption isotherms for CTF-1, PCTF-1, CTF-1-600, and PCTF-1-600 were determined, their uptake capacities at 273 K/1bar was stated to be in the order: FCTF-1-600 (243.6 mg/g) > FCTF-1 (206 mg/g) > CTF-1-600 (169 mg/g) > CTF-1 (108 mg/g). FCTF-1 showed the highest initial isosteric heat of adsorption (Qst) value of 35.0 kJ/mol at low coverage, demonstrating the moderate affinity to CO2. The CO2/N2 (10:90 v/v) selectivity established on singlegas adsorption isotherms at 298 K/1 bar, for these triazine-based COFs using ideal adsorption solution theory (IAST) (O'Brien & Myers, 1988) was also stated to decline as follows: 31 (FCTF-1) > 20 (CTF-1) > 19 (FCTF-1-600) > 13 (CTF-1-600). (Olajire, 2017). Other covalent triazine-based frameworks (CTFs) containing fluorene building blocks (fl-CTF300, fl-CTF350, fl-CTF400, fl-CTF500, fl-CTF600 have been synthesised by Lotsch and coworkers as shown in Scheme 13 (Hug et al., 2014) Scheme 13: Synthesis pathway for fl-CTFs. (Hug et al., 2014) However, the CTFs obtained at lower temperatures were reported mainly to possess ultramicropores and moderate surface areas of 297 m 2 /g at 300 °C. But the porosities of fl-CTFs materials obtained at higher temperatures were reported to increase significantly, giving rise to surface areas in excess of 2800 m 2 /g. (Hug et al., 2014) The COF fl-CTF-350 was reported to have a high CO2/N2 selectivity ratio of 37 as well as possess the highest fraction of micropores and moderate SBET of 1235 m 2 /g. However, the adsorption value for fl-CTF-350 (188 mg/g) is lower than the reported value for FCTF-1 (205-243 mg/g) (Olajire, 2017) (Bhunia, Boldog, et al., 2013;Bhunia, Vasylyeva, et al., 2013) and TPI-1-7 (29.9-107.8 mg/g). (Liebl & Senker, 2013) In addition, the CO2 adsorption capacity of fl-CTF-350 is higher than the uptake capacities of various Microporous Organic Polymers (MOPs) such as covalent organic frameworks (COFs; 53.2-169.8 mg/g), (Dawson et al., 2011;) microporous polyimides (MPIs; 99-167.6 mg/g)(G. Li & Wang, 2013) and hyper-cross-linked organic polymers (HCPs; 84-172.5 mg/g). (Luo et al., 2013;Martín et al., 2011). The synthesis of microporous covalent triazine-based organic polymers MCTP-1 and MCTP-2 was conducted by Ahn and coworkers (Puthiaraj et al., 2015) using cyanuric chloride with 1,3,5triphenylbenzene/trans-stilbene via Friedel-Craft reaction shown in Scheme 14.
Scheme 14: synthetic pathway for the MCTP networks. (Puthiaraj et al., 2015) These materials (MCTP-1 and MCTP-2) were reported to possess high SBETs of 1452 and 859 m 2 /g respectively. The reported CO2 adsorption capacity for MCTP-1 is 204.3 mg/g at 273 K/1 bar as well a moderate CO2/N2 selectivity ratio of 54.4. However, MCTP-2 was reported to show a moderate CO2 uptake of 160 mg/g at 273 K/1 bar with a moderately high CO2/N2 selectivity ratio of 68.6 compared to 54.4 for MCTP-1. In a bid to obtain new functionalized COF materials for gas capture, Hug et al. (Hug et al., 2015) synthesized nitrogen rich covalent triazine frameworks (CTFs) based on lutidine, pyrimidine, bipyridine and phenyl units in order to study their CO2 adsorption capacities. These materials were synthesized via ionothermal synthesis at different temperatures as shown in Scheme 15. These porous organic polymers (POPs) were reported to display high gas uptakes and exceptionally high CO2/N2 selectivity ratios. The CO2 adsorption capacity of bipy-CTF-600 (BP-CTF-600) was reported to be the highest in the family of CTFs synthesized, with CO2 uptake capacity of 245.5 mg/g at 278 K and CO2/N2 selectivity ratio of 20.1; while pyrimidine-based CTF-500 (Py-CTF-500) which contained the highest nitrogen content was reported to possess the highest CO2/N2 selectivity ratio of 50.2 based on IAST but with a lower CO2 capture capacity of 185 mg/g. In addition, Hug, et al. (2015) also reported that the trends in the gas adsorption behaviour within the CTF family and nitrogen-containing POPs, revealed the dominant role of the micropore volume for maximum CO2 uptake performance, while the Ncontent is a secondary factor for improving the CO2 adsorption capacity. However, the nitrogen content in the frameworks was stated to be a key contributor to the high CO2/N2 selectivity ratios of the produced CTFs. (Hug et al., 2015). Zhong and coworkers(Keke Wang et al., 2016) also designed and synthesized two covalent triazine-based frameworks with both ultramicropores and high nitrogen content for the purpose of CO2 capture. They successfully synthesized two covalent triazine-based frameworks (CTF-FUM and CTF-DCN) with ultramicropores of width less than 7 Å using fumaronitrile (FUM) and 1,4dicyanonaphthalene (DCN) as shown in Scheme 16. The CTF-HUST materials were reported to display good performance in the fields of photocatalyst, gas separations as well as in sodium ion batteries. (Kewei Wang et al., 2017) . Building units such as dicyano-and / or multicyano-, pyridine and thiadiazole compounds (Bojdys et al., 2010;Grill et al., 2007;Kuhn et al., 2009;Kuhn, Forget, et al., 2008) are the main monomers used for the construction of various 2-D CTFs. Some of these building units are shown in Figure 3. Triazine-based COFs show better thermal and chemical stability when compared to boron-based COFs, although they show poor crystallinity and porosity, as well as poor structural integrity. Irrespective of these shortcomings, CTFs show high CO2 uptake capacities and selectivity as well as being reported to show good reusability after several regenerations without loss of their CO2 adsorption abilities even in the presence of water vapour. (Olajire, 2017).

Imine-based COFs
Imine-based COFs were first investigated by Yaghi and coworkers(Uribe-romo et al., 2011) and are crystalline porous polymeric material containing C=N bonds. The Yaghi group applied the principles of dynamic covalent chemistry for the construction of imine based COFs. These COFs were of a ''Schiff base'' type formed by the co-condensation of aldehydes and amines or a ''hydrazine'' type that was made by the co-condensation of aldehydes and hydrazides. An imine-linked porous porphyrin polymer CuPor-BPDC was also synthesized by Neti et al. (Neti et al., 2013) using a Schiff base condensation reaction between 5, 10, 15, 20-terakis (p-aminophenyl) porphyrin Cu(II) and 4,4ˈ-biphenyl dicarboxaldehyde as shown in Scheme 20.
Scheme 24: The synthesis of ACOF-1.(Z. Li et al., 2014) ACOF-1 was reported to possess an SBET and pore size of 1176 m 2 /g and 0.94 nm respectively. In addition, it was reported that the CO2 uptake capacity of ACOF-1 was 177 mg/g at 273 K/1 bar as well as a CO2/N2 selectivity of 40 at 273 K in the very low pressure range of 0 -0.1 bar. Another 2-D azine-linked COF is COF-JLU2, reported by Li et al.(2015) also used a condensation reaction of hydrazine hydrate and 1,3,5triformylphloroglucinol under solvothermal conditions as shown in Scheme 25.
Scheme 25: The synthesis of COFJLU2. Li et al., (2015) COF-JLU2 contains many heteroatom sites in its skeleton, which are believed to enhance its functionality. COF-JLU2 was reported to be highly crystalline and to possess a moderate SBET and pore size of 410 m 2 /g at 77 K and 0.96 nm respectively as well as display good chemical (towards acid and base) and thermal stability. In addition, the CO2 adsorption capacity reported for COF-JLU2 was 216 mg/g at 273 K/1 bar. This CO2 uptake capacity value is higher than adsorption capacity values for some other COFs in the literature such as COF-103 (76.6 mg/g), ) TDCOF-5 (92 mg/g)  and ACOF-1 (177 mg/g).(Z. Li et al., 2014) Nevertheless, the uptake capacity of COF-JLU2 (216 mg/g at 273 K/1 bar) is comparable to uptake capacities of some amorphous COFs such as porous organic polymer CPOP-1 (212 mg/g)(Q. Chen et al., 2012), azo-linked polymer ALP-1 (267 mg/g) (Arab et al., 2014) and imine-linked porous polymer PPF-1 (269 mg/g).  The reported CO2 uptake capacity of COF-JLU2, was ascribed to the in-built micro-porosity of the network and the pore walls which are nitrogen-and oxygen-rich. The reported CO2/N2 selectivity ratio for COF-JLU2 was 77 which is higher than the CO2/N2 selectivity ratio for ACOF-1 (40)(Z. Li et al., 2014) and PPFs (21).  However, this same value (77) is less than CO2/N2 selectivity ratio for some microporous organic polymers (MOPs) such as tetraphenyladamantanebased polycyanurate network PCN-AD (112) (Shen et al., 2014) and benzimidazole-linked polymer BILP-2 (113). (Rabbani & El-Kaderi, 2012) In addition, the reported CO2/CH4 selectivity ratio for COF-JLU2 at temperatures of 273 K and 298 K were 4.1 and 3.2 respectively.(Z. Li et al., 2015) These two selectivity values are comparable to values reported in the literature for other microporous organic polymers (MOPs). (Arab et al., 2014;Q. Chen et al., 2012;Zhu et al., 2013). Another group of researchers, Smaldone and co-workers Alahakoon et al .(2016) also reported the synthesis of an azine-linked COF containing a sixfold symmetric hexaphenylbenzene (HEX) monomer functionalized with aldehyde groups which was thereafter treated with hydrazine to give HEX-COF1 as shown in Scheme 26.
Scheme 26: synthesis of HEX and HEX-COF1. Alahakoon et al. (2016) HEX-COF-1 was reported to have a mean pore size of 1 nm and a surface area in excess of 1200 m 2 /g as well as to demonstrate outstanding CO2 uptake capacity of 200 mg/g at 273 K and 1 atmosphere pressure. In addition, the design and construction of a laminar COF (RT-COF-1) from the reaction of 1,3,5-tris(4aminophenyl)benzene (TAPB) and 1,3,5benzenetricarbaldehyde (BTCA) through a Schiff base reaction at room temperature in air was conducted by Zamora and co-workers Delapeñaruigõmez et al. (2015) as shown in Scheme 27. The SBET and pore volume reported for RT-COF-1 at N2 isotherm of 77 K were 329 m 2 /g and 0.224 cm 3 /g respectively. (Delapeñaruigõmez et al., 2015) Furthermore, it was reported that activated RT-COF-1 adsorbed ca. 86 mg/g of CO2 at 273 K/1 bar. (Delapeñaruigõmez et al., 2015). Furthermore, Gao and co-workers  synthesized an azo-based COF, COF-Tpazo via a Schiff base condensation reaction of 4,4-azodianiline (Azo) and 1,3,5-triformylphloro-glucinol (Tp). COF-Tpazo was reported to have an SBET of 1552 m 2 /g and a pore volume of 0.97 cm 3 /g as well as a CO2 uptake capacity of 112 and 68 mg/g at 273 and 298 K/1 bar respectively. In addition, COF-Tpazo was also found to exhibit a high CO2/N2 selectivity ratio of 127 at 273 K; while the CO2/N2 selectivity ratio recorded at 298 K was 45. However, the CO2/CH4 selectivity ratios at 273 and 298 K, were reported to be 39 and 43 respectively. The CO2-philic and N2-phobic ends of COF-Tpazo were proposed to be responsible for the reported properties of the material.  Scheme 27: The room-temperature polyimine condensation to form RT-COF-1. Delapeñaruigõmez et al. (2015) Scheme 28: Synthesis of [N=N]x% -TAPH-COFs with azobenzene group and [C=C]x% -TAPH COFs with stilbene group via nucleophilic substitution reaction from [HO]x% -TAPH-COFs (X = 25, 50, 75, 100). Zhao et al. (2016) Furthermore, Gao and Co-workers also synthesised a range of tailored COFs , i.e [N=N]x%-TAPH-COFs and [C=C]x%-TAPH-COFs by post-modification of [OH]x%-TAPH-COFs with 4-phenylazobenzoyl chloride and stilbenecarbonyl chloride (PhStil),(S. Zhao et al., 2016) respectively as shown in Scheme 28. These COFs were reported to possess moderate surface areas, narrow pore sizes as well as good physicochemical stability. The [N=N]x% -TAPH-COFs were reported to show higher CO2 adsorption capacity of up to 207 mg/g at 273 K/1 bar and CO2/N2 selectivity ratio of 78 at 273 K than the corresponding [C=C]x% -TAPH-COFs (16-27).(S. Zhao et al., 2016) The differences in properties between these two groups of COFs were attributed to the dipole interaction between the azo group and the CO2 as well as the N2-phobic behaviour of the azo group. Three hypothetical 2-D squaraine-bridge covalent organic polymers (SQ-COP-1, SQ-COP-2 and SQ-COP-3) were proposed by Huang and Cao(L. Huang & Cao, 2016) by the use of linear squaraine unit and heterocyclic molecules such as (CH3)3B3O3, H3B3N3 and (CH3)3C3N3 as shown in Scheme 29 as well as the use of grand canonical Monte Carlo (GCMC) simulations.
Scheme 29: Synthesis pathway for three 2-D squaraine-bridge COFs. Huang and Cao (2016) These 2-D squaraine COFs (SQ-COP-1, SQ-COP-2 and SQ-COP-3) were calculated to exhibit exceptional high SBET of 8686, 8938 and 8585 m 2 /g respectively. The high SBETs associated with these 2-D squaraine COFs were attributed to the presence of the extended squaraine linkage in the framework materials. These exceptionally high SBETs are however, smaller than those of some microporous coordination polymers MCPs (10436 -10577 m 2 /g) (Schnobrich et al., 2010) but higher than those of MOF-210 (6240 m 2 /g), (Furukawa et al., 2010) PAF-1 (5460 m 2 /g) (Ben et al., 2009) and Nu-110SP (7800 m 2 /g). (Farha et al., 2012) The CO2 uptake capacities of SQ-COP-1, SQ-COP-2 and SQ-COP-3 were also estimated to be 990, 737 and 833 mg/g at 298 K/50 bar respectively. Nevertheless, it was reported that at 298 K /30 bar, the CO2 uptake capacities for SQ-COP-1, SQ-COP-2 and SQ-COP-3 were 804, 575 and 633 mg/g respectively which is significantly higher than the value 326 mg/g reported for a benchmark zeolite 13X. (Yuan et al., 2012) In addition, SQ-COPs COFs were reported to display a good selective adsorption of CO2 over N2, CH4, and H2. For example, the calculated CO2/N2 of SQ-COP-1 is comparable to that of PAF-302 (3-4)(Z. Yang et al., 2013) and AlBDC MOF (4.3).  Other imine-based COFs have also reported in the literature. Furukawa et al., 2016;Zou et al., 2017). The results described above show that imine-based COFs possess better crystallinity as well as structural regularity which can be tuned to achieve desired pore size more easily than triazine-based COFs (CTFs). In addition, imine-based COFs have better stability in water as well as in most organic solvents when compared with boron-based and triazine-based COFs. A summary of the main building units used for the construction of imine-based COFs is depicted below in Figure 4. The boron-, triazine-and imine-linked COFs discussed above make use of a conventional two-component [1 + 1] condensation design scheme. The design and synthesis of these COFs as well as new ones largely depends on the exploration of new nodes and linkers which can be difficult, time consuming and often do not give the desired results. In contrast to conventional two [1 + 1] design systems, multiple-component (MC) COF design makes use of three [1 + 2] i.e. one node and two linker units or four [1 + 3] i.e. one node and three linker units for their synthesis. Huang, Zhai, et al. (2016).

Multiple-component COFs
Multiple-component COFs (MC-COFs) feature asymmetric tiling of organic building units into anisotropic skeletons with uncommon shaped pores. Thus the application of the MC method has enabled the expansion of structural complexity of framework materials and their pores as well as the enhancement of their structural diversity. In addition, the multiplecomponent strategy has also provided new platforms for considerable expansion and design of functionalized structures of porous organic materials. The design and synthesis of both hexagonal and trigonal COF topologies using two-component [1 + 1] copolymerisation of a C3-or C4-symmetric node and a C2-symmetric linker Furukawa et al., 2016;Nagai et al., 2011;Ockwig et al., 2005c;San-Yuan Ding, 2013;Spitler et al., 2012) as shown in Figure 5 a, and three-or four-component systems Figure 5 b using one node and two or three dissimilar linkers.. Chen et al., 2015;Huang, Zhai, et al., 2016;Zeng et al.(2015) Figure 4: Some monomers used for constructing imine-based COFs. Adapted from refs. Chen et al., 2015;Huang, Zhai, et al.;Zeng et al. (2015) In a bid to enhance and expand the research of COF materials using multiple-component orthogonal reaction strategies, Zhao and co-workers used 4-formylphenyl boronic acid (a bifunctional linker) for the synthesis of MC-COF (NTU-COF-2). NTU-COF-2 was reported to consist of two dissimilar boronate and imine linkages in its skeleton that are not accessible for conventional [1 + 1] based COFs. (Zeng et al., 2015) By way of using different starting materials, Zhao and co-workers also designed a new three [1 + 2] multiple-component COF (NTU-COF-2) with the use of 2,3,6,7,10,11-hexahydroxytriphenylene (HHTP) which served as node while 4-formylphenylboronic acid (FPBA) and 1,3,5-tris(4-amino-phenyl)-benzene (TAPB) acted as the linkers to give an imine group and a C2O2B boronate ring as shown in Scheme 30.
Scheme 32: Synthesis and ordered structure of MC-COF-TP-E1E3E7 The same research(N. Huang, Zhai, et al., 2016) group used the MC/COF strategy with a [1+2] component reaction for the construction of tetragonal MC-COF (NiPc-E1E7) by the use of nickel phthalocyanine (NiPc) as the knot and two linkers (E1 and E7) as depicted in Scheme 33. It was reported that NiPc-E1E7 contained only one type of mesopore with pore size of 2.6 nm and SBET of 672 m 2 /g. The MC strategy of constructing COFs make use of a selective combination of three or more blocks (usually the building units used for the making of boron-, triazine-and imine-based COFs) consisting of one node and two or three linkers leading to the formation of hexagonal and tetragonal COFs. The MC strategy has opened new insight for the expansion and construction of structurally complex frameworks via asymmetric tiling of building units with provision for platforms for further construction of anisotropic πcolumnar arrays and unconventional shaped pores.

Synthetic Methods Used in the preparation of COFs:
In addition to the building units for the construction of COFs the synthesis conditions are also important to regulate the thermodynamic equilibrium involved in the formation of covalent bonds. Thus, the reaction media and conditions (temperature, pressure as well as presence or absence of template) are crucial factors for the formation of thermodynamically stable crystalline polymeric products. Since the pioneering work of Yaghi and co-workers, (Ockwig et al., 2005a) many other researchers globally have exploited and expanded the synthetic possibilities by developing mixed solvent systems and molten metal salts that have provided platforms such as solvothermal, ionothermal, microwaves and room temperature methods for COF synthesis.
Solvothermal Synthesis of COFs: Most of the reported COFs were obtained through solvothermal synthetic methods. A typical solvothermal method for synthesising COFs involves the mixing of monomers with a suitable solvent or a mixture of solvents and then putting the monomer mixture in a Pyrex or Parrautoclave followed by degassing with several freezepump-thaw cycles. Following the degassing process, the tube or the autoclave is sealed and heated to a given temperature (usually 80 °C and above) for some reaction time (usually 1-9 days).
Scheme 33: Synthesis of tetragonal [1+2] MC-NiPc-E1E7. Huang, Zhai, et al.. (2016) This treatment of monomers usually leads to the formation of precipitates. The precipitates are collected and washed with suitable solvent followed by drying under vacuum to give the COF material (usually a powder product). Factors such as solvent combination and ratios as well temperature and duration of reaction processes are crucial when designing a solvothermal reaction process for COFs synthesis, due to issues such as solubility, reaction rate, crystal nucleation and crystal growth as well as 'self-healing' of structures during the formation of COFs. Solvent combination and ratios are important points to consider for the purpose of forming crystalline framework materials. For example, the influence of solvents on the crystallinity of a COF materials was demonstrated by Jiang and coworkers  who found that cocondensation of zinc(II) 5,10,15,20-tetrakis(4-(dihydroxyboryl)phenyl)porphyrin and 1,2,4,5-tetrahydroxybenzene in a mixture of mesitylene and dioxane (1/1, v/v) produced amorphous solids. On the other hand, when the v/v ratio of mesitylene/dioxane was changed to 19/1 or 9/1, COF materials with high crystallinity were obtained. Furthermore, in the formation of boronate ester and boroxine-linked COFs, mesitylene/dioxane, (Colson & Dichtel, 2013;Ockwig et al., 2005a) DMAc-odichlorobenzene,  and THF/methanol (Tilford et al., 2006) have been used as solvent combinations. In the formation of borosilicate COFs, dioxane/toluene solvent mixtures have been used. (Hunt et al., 2008) Dioxane/aqueous acetic acid was used as solvent mixture in some imine-based COFs. (Uribe-romo et al., 2009) Mesitylene /dioxane/acetic acid solvent mixture (Kandambeth et al., 2015) was used as solvent mixture in the synthesis of some hydrazone-linked COFs. (Uribe-romo et al., 2011) Recently Zhao and co-workers used a mixture of mesitylene/dimethylacetamide/AcOH (6M, aqueous) (5:5:1 by volume) which was heated in a sealed glass ampoule at (120 °C for 3 days) to produce COF-BTA-DAB and COF-BTA-BZ. . A suitable temperature appropriate to the solvent used is also important to ensure the reversibility of the reaction. In general, most reported COFs in the literature were prepared in the temperature range 80 -120 °C depending on the chemical reactivity of the building units. A closed reaction system enhances the retention of any water molecules that could trigger a reverse reaction in the system. The final products of solvothermal methods are often powdered COF materials which may limit their practical applications under certain circumstance, e.g. interfaced incorporation into devices. To this end, Dichtel and co-workers (Colson et al., 2011) developed a solvothermal method for producing 2-D COF materials as thin films on single-layer graphene. It was reported that the obtained COF materials exhibited improved crystallinity in comparison with the powder samples. (Colson et al., 2011) Ionothermal Synthesis of COFs: Ionothermal synthesis involves the growing of single crystals from an ionic liquid in an autoclave at high temperature (300-600 °C) and pressure. Thomas and co-workers were the first to apply ionothermal method to produce crystalline porous COFs.  As mentioned in Section 2.3.2, they used reversible cyclotrimerisation of nitrile building units (1,4-dicyanobenzene) in molten ZnCl2 to generate covalent triazine-based frameworks (CTFs) with high crystallinity alongside outstanding chemical and thermal stabilities. The molten ZnCl2 functioned both as a solvent and as a catalyst for the reversible cyclotrimerisation reaction that lead to the formation CTFs. A comparison of COF materials obtained through solvothermal methods, and CTF polymers obtained from ionothermal method can be made; CTF polymers have the drawback of lack of crystallinity control because the reversible cyclotrimerisation reaction takes place under harsh reaction conditions that often lead to the formation of amorphous materials that lack long-range molecular order. (Jinang, 2012) In addition, the constraint of high reaction temperature limits the choice of suitable building units.(H. Ren et al., 2010) Based on these limitations, ionothermal methods at present have limited application.(San- Yuan Ding, 2013).
Microwave-assisted Synthesis of COFs: Microwave heating has many applications in accelerating chemical reactions. (Wei et al., 2015) Cooper and coworkers developed a rapid microwave-assisted protocol for the synthesis of boronate ester linked COFs. (Campbell et al., 2009) For example, 2-D (COF-5) and 3-D (COF-102) COFs could be synthesised using microwave heating in 20 minutes. (Wei et al., 2015) The syntheses of these COFs were achieved by conventional solvothermal methods in 72 h. (Ockwig et al., 2005a;Ritchie et al., 2010) Thus the microwave-assisted synthesis is 200 times faster than the solvothermal method. In addition, the SBET obtained for the product under microwave was reported to be 2019 m 2 /g; whereas the SBET of the same material but obtained via solvothermal means was reported to be 1590 m 2 /g. (Campbell et al., 2009) Thus by way of comparison, it can be seen that (1) Microwave-assisted synthesis can produce COFs more rapidly than solvothermal methods. (2) In microwaveassisted syntheses, the methods for the removal of solvent and residual impurities trapped in the frameworks are more efficient than in solvothermal methods, and thus microwave-assisted syntheses enhance the formation of materials with better porosity. (Jinang, 2012).
Challenges in the synthesis of COFs: Boron-, triazineand imine-based COFs that are obtained via condensation, Schiff-base reaction, Friedel-Crafts reactions, cyclotrimerization and post-synthesis exfoliation have all opened many opportunities for exploring novel functional materials using existing building units. (Beuerle & Gole, 2018;Diaz & Corma, 2016;San-Yuan Ding, 2013;Tong et al., 2014;Zeng et al., 2016;W. Zhao et al., 2018) However, some major challenges that are yet to be fully addressed are: further improvement on how to achieve the combined properties of stability and crystallinity as well as porosity in COF materials; the growing of extended and high quality crystalline molecular materials; and the design of COF materials with flexible monomers. Despite these challenges large SBETs and pore volumes as well as better structural integrity with low CO2 adsorption capacities at low pressures have been reported for boron-, boron/imine-, triazine and imine-based COFs. (Beuerle & Gole, 2018;Gao et al., 2018;Olajire, 2017) However, in terms of chemical/water stability as well diversity for different applications, further work is still needed. Ways forward to solving these challenges in COF synthesis are: (1) The development of new synthetic methodologies to improve the stability of COF materials, (2) The construction of COFs with more flexible monomers as well as diverse 3-D architectures for high-pressure CO2 uptake. (3) Development of room temperature methods for the synthesis of COFs.

Conclusion:
We have successfully reviewed the subject matter of COFs. The design methodologies and some applications of COFs and challenges associated with the making of COFs. Our anticipation is that this review article will help researchers to follow the global trend in order to achieve better molecular porous materials with ease.