Key words covalent organic frameworks - silver-catalyzed - three-component - one-pot - quinoline-linked
Introduction
Covalent organic frameworks (COFs) are an emerging class of crystalline porous materials
featuring periodic structures.[1 ]–[4 ] Owing to their designable structures and tailored functionalities, COFs have shown
great potential in broad fields including gas storage and separation,[5 ],[6 ] catalysis,[7 ]–[12 ] sensing,[13 ]–[17 ] drug delivery,[18 ]–[21 ] energy storage,[22 ]–[24 ] and so on. Since the first two COFs were reported by Yaghi in 2005,[25 ] progress in organic chemistry has shown significant contributions to the construction
of COFs,[26 ],[27 ] and their formation mechanisms have been deeply explored.[28 ]–[31 ] The formation of periodic COF frameworks is guided by the principles of dynamic
covalent chemistry, in which the final products are determined by their thermodynamic
equilibrium.[32 ],[33 ] Nevertheless, this intrinsic reversibility concurrently imparts inherent instability
to the resultant frameworks. To develop novel COFs and/or customize their functionality,
researchers have undertaken the assembly of periodic COF frameworks using irreversible
reactions, resulting in significant and noteworthy advancements. For instance, a variety
of synthetic strategies, including the direct synthesis method, reversible covalent
bond-based tandem reactions, post-synthetic modification, and multicomponent reactions,
are used for building polyimide COFs,[18 ],[34 ]–[37 ] olefin-linked COFs,[38 ]–[42 ] dioxin-linked COFs,[43 ]–[45 ] and many others.[46 ]–[51 ] As we all know, metal-catalyzed reaction chemistry, a crucially important part of
chemistry, has found extensive applications in organic synthesis, industrial production,
and biomedicine.[52 ],[53 ] At present, COF materials are mainly used as carriers of active metal centers to
carry out catalytic reactions of small molecules. It is noteworthy that there are
currently few reports regarding the synthesis of COF materials through metal-catalyzed
irreversible reactions.[54 ] This scarcity can be attributed primarily to the near insolubility of porous materials,
which significantly hampers the advancement of reactions and the expansion of frameworks.
Consequently, the utilization of metal-catalyzed reactions to achieve high crystallinity
and porosity COF materials with new structures and new functions is of great significance.
Herein, we introduce the metal-catalyzed reaction to synthesize COF materials.
This report was inspired by a silver-catalyzed three-component approach to quinolines
starting from anilines, aldehydes, and alcohols ([Scheme 1a ]).[55 ] Building upon this foundation, we have advanced a silver-catalyzed three-component
one-pot approach, including Schiffʼs base reaction to construct the COF backbone followed
by the key silver-catalyzed sequential process for the formation of two carbon–carbon
(C–C) bonds, gradually constructing COF materials based on multi-substituted quinolines
(Figure S1). Specifically, our investigation focuses on the reaction involving 1,3,5-tris(4-aminophenyl)benzene
(TAPB) paired with 2,5-dimethoxyterephthalaldehyde (DMTP), alongside various alcohols
(namely, ethylene glycol, ethyl lactate ester, and L-diethyl malate), in the presence
of silver trifluoromethanesulfonate (AgOTf) and trifluoromethanesulfonic acid (HOTf)
to yield polysubstituted quinoline-linked periodic frameworks ([Scheme 1b ]). The silver-catalyzed three-component reaction was performed in mesitylene, sealed
in an atmospheric atmosphere, and then heated at 120 °C without disturbance for 24 h.
Crystalline solids of quinoline-linked COFs (Q-COF-1, -2, and -3) were obtained with
isolated yields of 98%,[56 ] 56%, and 61%, respectively (see the Supporting Information for details).
Scheme 1 a) Synthesis of model molecule. b) Synthesis of Q-COF-1, -2, and -3.
Results and Discussion
To confirm the structure of the obtained materials, we synthesized COF-1 in mesitylene
for comparison (see the Supporting Information for details).[57 ] The formation of quinoline-linked Q-COFs (Q-COF-1, -2, and -3) was initially assessed
by Fourier transform infrared spectroscopy. As shown in Figures S2 – S4, the disappearance
of C(=O)–H (2869 and 2759 cm−1 of DMTP), N−H (3496 – 3326 cm−1 of TAPB) and O–H (3420 – 3345 cm−1 ) vibration indicated a high degree of polymerization by consuming almost all the
aldehyde, amine, and alcohol groups of the monomers. In addition, the appearance of
absorption bands around 1615 cm−1 suggests the formation of C=N bonds by a Schiffʼs base reaction, together with a
weak peak around 1642, 1558, and 1238 cm−1 , indicating the existence of a quinoline moiety (Figure S5). Furthermore, the appearance
of absorption bands of C (OMe)=O at 1676 and 1678 cm−1 further supports the formation of a quinoline moiety (Figures S3 and S4). Solid-state
13 C cross-polarization magic-angle-spinning nuclear magnetic resonance (CP/MAS NMR)
spectra of Q-COFs further supports the formation of quinoline units (Figures S6 – S8).
The existence of carbon peaks at 154, 154, and 153 ppm was observed, which could be
attributed to C=N bonds of quinoline for Q-COF-1, Q-COF-2, and Q-COF-3, respectively.
In contrast to Q-COF-1 (Figure S6), the presence of distinct carbon peaks at 180,
53, and 18 ppm, associated with C (O)CH2 CH3 , C(O)C H2 CH3 , and C(O)CH2
C H3 respectively, substantiates the synthesis of functionalized Q-COF-2 through the utilization
of ethyl lactate ester (Figure S7). For Q-COF-3, comparable carbon peaks resembling
those found in Q-COF-2 were identified. Nonetheless, a notable divergence is observed
with the emergence of an enveloped peak at 55 ppm, attributed to the C(O)C H2 CH3 and OC H3 functionalities originating from L-diethyl malate and DMTP, respectively. Additionally,
there is a discernible enhancement in the methyl signal corresponding to C(O)CH2
C H3 (Figure S8). X-ray photoelectron spectroscopy analyses provided additional validation
for the above results. Specifically, the N1s peak was detected at 401.13, 401.22,
and 400.99 eV, corresponding to the N1s of C=N bonds within quinoline moieties.[49 ],[58 ] This observation unequivocally confirms the successful formation of quinoline units
(Figures S9–S11). Field emission scanning electron microscopy and transmission electron
microscopy show that Q-COF-1 exhibits a rod-shaped morphology (Figures S12 and S13),
Q-COF-2 adopts a nanoparticle morphology (Figures S14 and S15), and Q-COF-3 displays
a flake aggregation morphology (Figures S16 and S17).
In addition, the crystal structures of these samples were elucidated utilizing powder
X-ray diffraction (PXRD), employing comparison with optimized, idealized structural
models of the expected frameworks. As anticipated, the diffraction patterns of the
three structurally analogous frameworks exhibit comparable characteristic reflections
([Figure 1a ]). The PXRD pattern, as depicted in [Figure 1 ], illustrates a series of prominent peaks at 2θ = 2.67° (100), 4.80° (110), 5.56° (200), and 7.40° (210) for Q-COF-1; 2θ = 2.68° (100), 4.72° (110), 5.47° (200), and 7.27° (210) for Q-COF-2; and 2θ = 2.67° (100), 4.74° (110), 5.46° (200), and 7.29° (210) for Q-COF-3. Lattice modeling
and Pawley refinement were conducted using Materials Studio software to generate their
probable structures characterized by 2D AA stacking and AB stacking. The comparison
between the experimental PXRD patterns and the simulated ones indicates a closer alignment
of the experimental diffraction peaks ([Figure 1 ], black) with the simulated patterns exhibiting AA stacking ([Figure 1 ], orange) better than those featuring AB stacking ([Figure 1 ], purple). The difference plots ([Figure 1 ], gray) suggest that the refined PXRD patterns ([Figure 1 ], red dot) are consistent with the experimental ones. Pawley refinements were conducted
to determine the unit cell parameters. For Q-COF-1, the refined parameters were found
to be a = b = 37.07 Å, c = 3.94 Å, α = β = 90°, and γ = 120°, yielding residuals R
p = 4.26% and R
wp = 5.33% (Tables S1 and S4). Similarly, for Q-COF-2, the parameters were determined
as a = b = 36.97 Å, c = 4.10 Å, α = β = 90°, and γ = 120°, resulting in residuals R
p = 5.33% and R
wp = 7.03% (Tables S2 and S5). Lastly, for Q-COF-3, the refined parameters were a = b = 36.93 Å, c = 4.82 Å, α = β = 90°, and γ = 120°, with residuals R
p = 3.59% and R
wp = 4.48% (Tables S3 and S6). These values closely approximate the corresponding structural
models, further confirming that the successful synthesis of Q-COFs adopts AA stacking
consistent with COF-1.
Figure 1 a) Experimental PXRD patterns of COF-1, Q-COF-1, Q-COF-2, and Q-COF-3. Experimental
(black), Pawley-refined (red) PXRD patterns, difference plot between the observed
and refined patterns (light grey), AA stacking patterns (orange), and AB stacking
patterns (purple) for b) Q-COF-1, c) Q-COF-2, and d) Q-COF-3.
To assess their permanent porosity, nitrogen sorption measurements of COF-1 and Q-COFs
were further carried out at 77 K. As shown in [Figure 2 ], This sorption profile of Q-COFs and COF-1 is best described as a type IV isotherm
with rapid N2 uptakes at the low relative pressure range P /P
0 < 0.05, which is characteristic of mesoporous materials. Their Brunauer−Emmett−Teller
(BET) surface areas were calculated to be 2200, 1123, 1261, and 1245 m2 /g (Figure S18), with total pore volumes (at P /P
0 = 0.99) being 2.116, 0.855, 0.812, and 0.933 cm3 /g for COF-1, Q-COF-1, Q-COF-2, and Q-COF-3, respectively. In comparison, the BET
surface areas and pore volume of the Q-COFs synthesized via the silver-catalyzed multicomponent
one-pot reaction are notably lower than those of the reference COF-1. Their pore size
distributions (PSDs) of Q-COFs and COF-1 calculated by quenched solid functional theory
were evaluated to be 3.43 nm for COF-1, 3.30 nm for Q-COF-1, and 2.89 nm for Q-COF-2
and Q-COF-3, respectively (Figure S19). For COF-1 and Q-COF-1, their PSD corresponded
well with previously reported findings.[57 ] In the case of Q-COF-2 and Q-COF-3, their PSD aligned closely with the theoretical
pore diameter associated with their eclipsed (AA) layer stacking configuration (2.52 nm
for Q-COF-2 and 2.41 nm for Q-COF-3, Figure S20). Other minor peak distributions indicate
the existence of defects in the obtained COFs.
Figure 2 N2 sorption isotherm curves of a) COF-1, b) Q-COF-1, c) Q-COF-2, and d) Q-COF-3.
The thermal stability of Q-COFs was assessed utilizing a thermogravimetric analyzer
under a nitrogen atmosphere. As illustrated in Figure S21, for Q-COF-1, the temperature
reached approximately 252 °C, resulting in a thermal weight loss proportion of approximately
1.5%, with the temperature corresponding to a 5% thermal weight loss recorded at 313 °C.
Similarly, for Q-COF-2, the temperature reached around 264 °C, accompanied by a thermal
weight loss proportion of approximately 1.5%, and a corresponding temperature for
a 5% thermal weight loss recorded at 343 °C. For Q-COF-3, the temperature reached
approximately 262 °C, resulting in a thermal weight loss proportion of approximately
1.5%, with the temperature corresponding to a 5% thermal weight loss recorded at 307 °C.
These findings indicate the favorable thermal stability properties of Q-COFs.
Conclusions
In summary, we have presented an efficient method for fabricating stable, crystalline,
porous quinoline-linked COFs. This multi-component reaction involves readily available
aldehydes, amines, and alcohols in a one-pot process via a silver-catalyzed cyclization
reaction. Furthermore, the functionalization of frameworks is achieved through the
direct substitution of alcohols. We anticipate that this strategy will provide a versatile
and feasible method for constructing numerous novel COF materials and will be adopted
by others and used to explore further material applications, possibly greatly expanding
the family of COF materials.
Funding Information
This work was supported in part by the National Natural Science Foundation of China
(Grant Nos. 21 922 502, 22 171 022, 21 971 017), the Natural Science Foundation of
Shandong Province (Grant No. ZR2021QB170), China Postdoctoral Science Foundation (Grant
No. 2021M700 416), and Beijing Municipal Science and Technology Commission (No. Z211100002421013).
