Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

# Two-dimensional semiconducting covalent organic frameworks via condensation at arylmethyl carbon atoms

## Abstract

Construction of organic semiconducting materials with in-plane π-conjugated structures and robustness through carbon-carbon bond linkages, alternatively as organic graphene analogs, is extremely desired for powerfully optoelectrical conversion. However, the poor reversibility for sp2 carbon bond forming reactions makes them unavailable for building high crystalline well-defined organic structures through a self-healing process, such as covalent organic frameworks (COFs). Here we report a scalable solution-processing approach to synthesize a family of two-dimensional (2D) COFs with trans-disubstituted C = C linkages via condensation reaction at arylmethyl carbon atoms on the basis of 3,5-dicyano-2,4,6-trimethylpyridine and linear/trigonal aldehyde (i.e., 4,4″-diformyl-p-terphenyl, 4,4′-diformyl-1,1′-biphenyl, or 1,3,5-tris(4-formylphenyl)benzene) monomers. Such sp2 carbon-jointed-pyridinyl frameworks, featuring crystalline honeycomb-like structures with high surface areas, enable driving two half-reactions of water splitting separately under visible light irradiation, comparable to graphitic carbon nitride (g-C3N4) derivatives.

## Introduction

Recently, 2D organic porous materials with well-defined π-extended structures, emerging as organic graphene analogs, have been successfully synthesized by connecting organic building blocks through carbon-carbon bond linkages1,2,3. They inherit the structural merits of graphene and π-conjugated organic molecules to some extent, promising for applications in optoelectronic and energy-storage devices4,5. Among these materials, a nano-porous graphene and a 2D conjugated aromatic polymer with carbon-carbon single bond (C – C) linkages have been synthesized through surface-mediated polymerization and solid-state topochemical polymerization, respectively6,7. In studies on these materials, rationally designed monomers for preassembly and precise regioselective coupling polymerization act as key roles in the formation of architectures with excellent topologies. In another case, we reported a 2D cyanostilbene-based covalent organic framework (COF) involving the connection of aromatic units through a carbon-carbon double bond (C = C) linkage under the Knoevenagel reaction (Fig. 1a)8. This report suggested that carbon-carbon bonding enables a system to undergo a thermodynamically controlled process, in which the crystallinity of an in-plane structure is governed through continued 2D chain growth and defect correction, as verified in the previously reported COFs with dynamic covalent linkages (e.g., imine)9. Afterwards, another 2D cyanostilbene-based COF was developed through a similar preparation protocol by using pyrene as one of the building blocks (Fig. 1a)10. Such kind of COFs exhibit π-delocalization over 2D networks, attributed to the efficient π-electron communication through C = C linkages, as verified by numerous C = C-containing artificial organic compounds and natural products (e.g., β-carotene) with unique photophysical properties. In a Knoevenagel reaction, the aryl α-carbon atom in a monomer can be easily converted to a reactive intermediate carbanion, which tends to be stabilized through a p-π-conjugated effect11. This carbanion can definitely promote the formation of a highly crystalline structure through self-healing processes. On the other hand, a C = C linkage enables avoiding the twisted conformations arising from steric repulsion in a main backbone when two aromatic units are directly coupled by a C – C single bond, thus beneficial to the formation of an in-plane π-conjugated structure. Nevertheless, it would be challenging to determine appropriate conditions or regulation for reversible C = C bond formation, even in the aforementioned Knoevenagel reactions, which, according to our knowledge, only succeeded for a few COF samples arising from 1,4-phenylenediacetonitrile (or its derivatives).

Owing to the highly exposed surface areas and tunable band gaps, polymeric semiconductors such as graphitic carbon nitride (g-C3N4) have been attracting considerable attention as high-performance photocatalysts for water splitting under visible light, which is one of the long-term goals for the efficient storage of intermittent solar energy12,13,14,15,16,17,18,19,20. The high crystallinity of 2D COFs has substantial advantages with respect to their well-defined structures, which can facilitate establishing structure-property relationships and provide insights into photocatalytic processes or mechanisms21,22,23,24. Attempts to promote the photocatalytic activities for COFs have mainly been hindered by their relatively low stability and poor π-electron delocalization25,26. Pyridine derivatives as nitrogen-embedded aromatic units feature Lewis basicity and versatile optoelectronic properties. In particular, completely different from the interrupted conjugation effect on the meta-positions of phenyl ring, the facile modification of pyridine in its 2,6-positions, 2,5-positions, or 2,4-positions enables achieving well extended π-conjugated systems, thus extensively used as building blocks or ligands for constructing organic or organometallic semiconductors with excellent photocatalytic behaviors27,28.

In this study, we utilize 3,5-dicyano-2,4,6-trimethylpyridine (DCTMP) as the key building block, whose three methyl carbon atoms are readily converted to relatively stable carbanions under an organic base catalyst in a p-π-conjugated system29,30. Upon condensation with 4,4″-diformyl-p-terphenyl (DFPTP), 4,4′-diformyl-1,1′-biphenyl (DFBP) and 1,3,5-tris(4-formylphenyl)benzene (TFPB) in N,N-dimethylformamide (DMF), three 2D COFs (denoted as g-CxNy-COFs) with trans-disubstituted C = C bond linkages are formed in nearly quantitative yields, termed as g-C40N3-COF, g-C31N3-COF, and g-C37N3-COF, respectively. Powder X-ray diffraction (PXRD), high-resolution transmission electron microscopy (HRTEM), and surface area measurement clearly reveal the COFs have honeycomb-like crystalline porous structure with high surface areas. The appropriate energy levels of conduction and valence bands of g-C40N3-COF allow for driving two half-reactions of water splitting separately to generate hydrogen or oxygen under visible light irradiation.

## Results

### Design and synthesis of g-CxNy-COFs

The target COFs were solvothermally synthesized through Knoevenagel condensation reaction between 3,5-dicyano-2,4,6-trimethylpyridine (DCTMP) and linear/trigonal aldehyde (4,4″-diformyl-p-terphenyl, 4,4′-diformyl-1,1′-biphenyl or 1,3,5-tris(4-formylphenyl)benzene) in DMF, catalyzed by the organic base piperidine. In this process, first, the pyridinyl methyl carbon atoms in DCTMP are converted to carbanions through C–H cleavage upon treatment with the base. This conversion is typically attributed to a decline in the electronic cloud density around these arylmethyl carbon atoms, which is caused by the strong electron-deficient pyridine ring that bears two electron-withdrawing cyano groups. The resulted carbanions tend to be stabilized through a strong p-π-conjugated interaction with this substituted pyridine ring, enabling the subsequent nucleophilic attacks on the positive carbon atoms in the aromatic formyl groups of these aldehydes. Such attacks form carbon-carbon bonds, which then transform to carbon–carbon double bonds by releasing H2O molecules. Under the guidance of such a reaction mechanism, we established an appropriate condition for achieving high-crystalline olefin-linked COFs. For example, in an optimized procedure, DCTMP (0.50 mmol), DFPTP (0.75 mmol), and piperidine (3.00 mmol) were dissolved in DMF (10 mL) and employed in a 15-mL pressure flask. The mixture was heated at 150 °C for 72 h, after which the resulted yellow solid was collected and purified through a standard workup to obtain the target COF g-C40N3-COF with nearly quantitative yield (see details in the methods section). In order for comparison, a model compound 3,5-dicyano-2,4,6-tristyrylpyridine (denoted as DCTSP) was also synthesized by a reaction of DCTMP and benzaldehyde under the same reaction condition.

### Characterizations of g-CxNy-COFs

PXRD analyses revealed the crystallinity of the resulting COFs. In conjunction with structural simulation, the structural features of g-CxNy-COFs were elucidated. Of these, the obtained PXRD pattern of g-C40N3-COF was dominated by an intense reflection in the low-angle region at 2θ = 2.62°, which was assigned to the (100) facet (Fig. 2b). Four other peaks at 4.53°, 5.23°, 6.94°, and 26.67° were assigned to the (110), (200), (210), and (001) facets, respectively (Supplementary Fig. 14). Fully eclipsed AA layer stacking and staggered AB layer stacking models were generated. Geometrical energy minimizations of the structural models were conducted using the Materials Studio software package and diffraction patterns were simulated. The experimental PXRD pattern agreed well with a pattern simulated from an AA-eclipsed layer stacking model with unit cell parameters (a = 37.419 Å, b = 37.843 Å, c = 3.588 Å and α = 90.044°, β = 90.061°, γ = 119.756°) in the space group $${\it{P}}\bar 1$$. While, the staggered AB layer stacking model did not match the experimental pattern. Finally, Pawley refinement against experimental PXRD data provided good agreement factors (Rwp = 3.47% and Rp = 2.75%). In addition, the (100) peak had a small full-width at half-maximum value of 0.36°, which is comparable to those of imine-based COFs, indicating the high crystallinity of the framework. As contrast, g-C31N3-COF and g-C37N3-COF exhibited lower crystallinity, as demonstrated by their broadened diffraction peaks of (100) facets at 2θ = 3.36° and 5.61°. Simulated AA stacking modes matched better with their PXRD patterns, from which hexagonal pore diameters were estimated as 2.6 nm and 1.6 nm for g-C31N3-COF and g-C37N3-COF, respectively (Fig. 2c, d).

The permanent porosities of g-CxNy-COFs were investigated by nitrogen sorption analysis at 77 K. As illustrated in Fig. 2e, f, the nitrogen sorption isotherms of g-C40N3-COF and g-C31N3-COF are assigned to type-IV reversible isotherms, indicative of their mesoporous characteristics. Whereas, g-C37N3-COF shows type-I reversible isotherm, suggesting its microporous structure (Fig. 2g). The Brunauer-Emmett-Teller (BET) surface areas of g-C40N3-COF, g-C31N3-COF, and g-C37N3-COF were calculated to be 1235 m2 g-1, 864 m2 g-1, and 1012 m2 g-1, respectively, which were much higher than those for the other C = C linked COFs reported to date (Fig. 1a)8,10. The pore size distribution (PSD) was evaluated with non-local density functional theory (NLDFT). The PSDs of g-CxNy-COFs exhibited maxima of 3.2 nm, 2.43 nm, and 1.57 nm, which were highly in line with the predicted values for AA eclipsed geometries of the frameworks.

The defined chemical structures of g-CxNy-COFs were confirmed through 13C cross-polarization/magic angle spinning solid-state nuclear magnetic resonance (CP/MAS ssNMR) spectroscopy, Fourier transform infrared (FT-IR) spectroscopy, thermogravimetric analysis (TGA), and elemental analysis. As an example, 13C ssNMR spectra of g-C40N3-COF and 13C NMR spectra of the model compound DCTSP in CDCl3 were combined in Fig. 3a. The peak at 117 ppm was assigned to the carbon atoms in cyano groups (a). The carbon atoms on pyridine unit (b, c, and d) corresponded to the peaks at 103, 159, and 155 ppm, respectively. The signals of vinyl carbon atoms (e, f, h and i) were found at 119, 120, 135, and 136 ppm, respectively. The intense resonance at 127 ppm was ascribed to the hydrogen-bonded aromatic carbons (g). The peak at 140 ppm was attributed to the carbon-attached aromatic carbons (j). FT-IR spectra (Fig. 3b) revealed that the stretching vibration peak of the cyano group (CN) was located at approximately 2225 cm−1, which was also observed for the monomer DCTMP. Moreover, a peak which was observed at 1690 cm−1 attributed to the C=O stretching vibration of the monomer DFPTP, disappeared in g-C40N3-COF, indicating a high polymerization degree. Notably, in addition to the peak at 1621 cm−1, which was attributable to the C = C moiety, the peak at 970 cm−1 engendered by the trans-HC = CH stretch vibration clearly manifested the trans-configurations of disubstituted olefin linkages in the entire framework. TGA revealed the extremely high thermal stability of g-C40N3-COF in N2 up to 500 °C, with the framework exhibiting less than 10% weight loss (Supplementary Fig. 6). Elemental analysis showed that the elemental contents (C: 84.68%, N: 8.38% and H: 4.95%) were close to the theoretical data (C: 87.89%, N: 7.69%, and H: 4.43%). X-ray photoelectron spectroscopy (XPS) further provided detailed information regarding the chemical structure of the framework (Supplementary Fig. 4). In high-resolution spectra, the C 1 s peak at 284.5 eV and the N 1 s peak at 399.0 eV were assigned to the sp hybrid C≡N moiety in cyano group. The peaks at 286.3 and 285.0 eV in the C 1 s zone were attributed to electrons originating from sp2 carbon atoms in vinyl groups and aromatic rings, respectively. The peak at 399.5 eV in the N 1 s zone was ascribed to pyridine nitrogen atoms. Similarly, the structural characterizations of g-C31N3-COF and g-C37N3-COF were also systematically performed for the verification of their chemical structures (Supplementary Figs. 16).

HRTEM was used to further confirm the formation of a periodic porous framework for g-C40N3-COF. The lattice fringes are depicted in Fig. 4a, revealing an ordered alignment with a high degree of crystallinity. A honeycomb-like porous structure was clearly visualized along the [001] plane under high resolution (Fig. 4b). As indicated by a fast Fourier transform (inset in Fig. 4a) conducted on a selected area (red square in Fig. 4a), a hexagonal arrangement of six white diffraction spots was found in the raw HRTEM image, corresponding to (100) facet. An observation from the [110] direction (Fig. 4d) revealed the distinct one-dimensional channels with a uniform diameter of ~3.28 nm (Fig. 4e), which is consistent with the pore diameter of 3.3 nm in the simulated COF structure, according to PXRD analysis. In most cases of COFs, limited HRTEM images were achieved, primarily because skeletal structures could be easily damaged under high-energy electron beams. Recently, a low-dose TEM technique was developed for realizing well-resolved images of COFs with electron beam-sensitive linkages (e.g., imine bonds)31. By contrast, g-C40N3-COF could be clearly visualized even under normal TEM measurement conditions without any additional techniques. This phenomenon strongly supports the robust network of the as-synthesized COF through C = C linkages. Clear observation of ordered domains in the TEM images of g-C31N3-COF and g-C37N3-COF was difficult, likely due to their lower crystallinity than g-C40N3-COF (Supplementary Fig. 17).

### Photocatalytic properties

To gain more insights on the active sites of g-C40N3-COF for H2 evolution, the reaction process of proton-adsorption-reduction-hydrogen-adsorption was simulated using DFT calculations and the free-energy changes were calculated regarding to two types of nitrogen atom and six carbon atoms of g-C40N3-COF (Fig. 6e). The free-energy variations indicate that site N2 (pyridine-N) is favorable for single-site H2 evolution, implying that nitrogen atoms of pyridinyl rings in the network of g-C40N3-COF are the dominant active sites for photocatalytic H2 evolution, well in agreement with our previous discussion.

The other two COFs, g-C31N3-COF, and g-C37N3-COF were also investigated as photocatalysts for H2 evolution under the same reaction conditions. The preliminary results demonstrate that the H2 evolution rates for g-C31N3-COF and g-C37N3-COF are 27.1 and 19.8 μmol h−1, respectively, which are much lower than that of g-C40N3-COF (Supplementary Fig. 31). Apart from the light-harvesting and charge separation abilities, the crystallinity for these COF photocatalysts also serve as the crucial role in their photocatalytic activities for water-splitting, likely associated with the defects over the whole networks, as demonstrated by Cooper et al. very recently37.

Clearly, achieving an overall water-splitting process using artificial photocatalysts is the ultimate goal in the pursuit of renewable energy sources. Nevertheless, the water oxidation process can be sluggish, presumably due to low driving force and slow O-O bond formation kinetics, as observed for the representative metal-free photocatalyst g-C3N439,40. In this study, the water oxidation performance of g-C40N3-COF was evaluated using AgNO3 as an electron acceptor and La2O3 as a pH buffer under visible-light irradiation. The unmodified g-C40N3-COF sample demonstrated a relatively low O2 production rate of 1.6 μmol h−1 (for 50 mg g-C40N3-COF). Loading cobalt species instead of noble metals into the COF samples as cocatalysts could engender a distinct performance enhancement. A low-cost non-noble metal salt Co(NO3)2, as the cobalt source (3 wt% Co2+), was introduced into the reaction system to modify the catalyst. As illustrated in Fig. 6d, the modified catalyst exhibited visible-light activity (λ > 420 nm) for O2 evolution with a rate of 2.5 μmol h−1. The resulting oxygen evolution rate is even better than that of bulk g-C3N412. Although the O2 evolution activity of g-C40N3-COF was determined to be low, these results directly manifested the thermodynamic possibility of oxygen evolution as a reality. The low activity for O2 evolution might be ascribed to the smaller thermodynamic driving force of such type of COF as compared with that of H2 evolution, which could be further improved by rationally tuning semiconducting properties through precise structural modifications.

## Discussion

Typically, 2D COFs produced through C = C linkages possess unique geometric and electronic characteristics because they inherit notable advantages from conventional linear conjugated polymers and 2D graphene. The intriguing physical properties of these types of polymers render them highly desirable as semiconductors for optoelectronic applications. The traditional concept is that carbon-carbon bonds are irreversible and thus could not be used in the construction of highly crystalline structures through a thermodynamic process. Three years ago, we reported an example of a crystalline COF with cyano-substituted C = C linkages, which revealed the reversibility of C = C bonds in a Knoevenagel condensation reaction, allowing the formation of highly ordered structures through self-healing processes. In this study, we successfully constructed another family of 2D COFs linked by trans-disubstituted C = C bonds through condensation at arylmethyl carbon atoms. The highly crystalline frameworks were clearly revealed by PXRD, HRTEM, and surface area measurements. The frameworks’ intrinsic chemical structures, comprising pyridine core networks and neat C = C linkages, grant them π-electron delocalized properties and excellent light-harvesting characteristics. In addition, the appropriate band energy levels render the one of the framework g-C40N3-COF suitable as a photocatalyst for water splitting with two half-reactions in the presence of sacrificial reagents. A high hydrogen production rate was achieved, namely 206 µmol h−1 (relative to 50 mg of COF sample), with the apparent quantum efficiency of 4.84% at λ = 420 nm. This is among the highest values reported for COF-based photocatalysts so far. An oxygen evolution rate of 2.5 μmol h−1 was detected for COF photocatalyst when silver nitrate was applied as an electron acceptor under visible light, comparable to bulk g-C3N4. One can reasonably foresee that this work provides an approach to scalable and sustainable preparation of excellent 2D organic semiconductors or organic graphene analogs for potential practical applications far beyond photocatalysis.

## Methods

### Synthesis of g-CxNy-COF

All Chemicals were used as received without further purification. First, 0.50 mmol (85.60 mg) of 3,5-dicyano-2,4,6-trimethylpyridine (DCTMP) and 0.75 mmol (214.74 mg) of 4,4″-diformyl-p-terphenyl (DFPTP) (or 0.75 mmol (157.67 mg) of 4,4′-diformyl-1,1′-biphenyl (DFBP) or 0.50 mmol (195.22 mg) of 1,3,5-tris(4-formylphenyl)benzene (TFPB)) were placed in a 15-mL pressure flask. Then it was transferred into an argon-filled glovebox, where 3.00 mmol (255.45 mg) piperidine and 10 mL anhydrous deoxygenated N,N-dimethylformamide (DMF) were added. The mixture was stirred in glovebox for 5 min to dissolve the solid monomers into DMF. Then the pressure flask was sealed by Teflon nut and taken out from glovebox. Subsequently, the clear homogenous pale-yellow solution was heated in a 150 °C oil bath for 3 days. After cooling to room temperature, yellow precipitates were observed at the bottom of flask, leaving clear and colorless supernatant. The precipitate was filtered and washed by acetone and dichloromethane for several times, and then it was dried under vacuum for 12 h at 120 °C. Finally, yellow powder was obtained and the yield based on monomers was calculated to be approximately 99.3%.

### Photocatalytic hydrogen evolution

A flask was charged with 50 mg of g-C40N3-COF powder, 100 mL deionized water and 10 mL triethanolamine (TEoA). It was ultrasonicated for 15 min to obtain a well-dispersed suspension. Then the resulting suspension was transferred into a Pyrex top-irradiation reaction vessel connected to a closed gas system. Certain amount of platinum (Pt) as cocatalyst was loaded into the network of photocatalyst by in situ photodeposition method using H2PtCl6. The reaction mixture was evacuated several times to ensure complete removal of air prior to irradiation in a 90° angle with a 300 W Xe light-source. The wavelength of the incident light was controlled by using a 420 nm long pass cut-off filter. The temperature of the reaction solution was maintained at 12 °C by the flow of cooling water. The evolved gases were analyzed by gas chromatography equipped with a 5 Å molecular sieve column at 60 °C with argon as the carrier gas. Hydrogen was detected with a thermal conductivity detector (TCD) referencing against standard gas with a known concentration of hydrogen. Hydrogen dissolved in the reaction mixture was not measured and the pressure increase generated by the evolved hydrogen was neglected in the calculations. The hydrogen evolution rates were determined from a linear regression fit.

### Photocatalytic oxygen evolution

Fifty miligram of g-C40N3-COF was well dispersed by ultrasonication in an aqueous solution (100 mL) containing 0.01 mol L−1 AgNO3 as an electron acceptor and 0.2 g La2O3 as a pH buffer. The suspension was poured into a Pyrex top-irradiation reaction vessel connected to a closed gas system. Then it was evacuated several times to completely remove air prior to irradiation under a 300 W Xe lamp. The temperature of the reaction solution was maintained at 12 °C by the flow of cooling water. The evolved gases were analyzed by gas chromatography equipped with a 5 Å molecular sieve column at 60 °C with argon as the carrier gas. Oxygen was detected with a thermal conductivity detector (TCD) referencing against standard gas with a known concentration of oxygen.

### The apparent quantum yield (AQY) measurements

The apparent quantum yield (AQY) for H2 evolution was measured using monochromatic LED lamps with band pass filter of 420 ± 4.6 nm, 470 ± 4.4 nm, 490 ± 3.8 nm, 520 ± 4.0 nm, 578 ± 4.1 nm (errors of wavelength were depended on Guass Simulation of full width at half maximum), the intensities were 14.3, 11.2, 7.6, 2.9, 1.6 mW cm−1, respectively (ILT 950 spectroradiometer). The irradiation area was controlled as 3 × 3 cm2. Depending on the amount of hydrogen produced by the photocatalytic reaction in one hour, the AQY was calculated according to Eq. (1):

$$\begin{array}{c}{\mathrm{\eta }}_{{\mathrm{AQY}}} = \frac{{{\mathrm{N}}_{\mathrm{e}}}}{{{\mathrm{N}}_{\mathrm{p}}}} \times 100\% \\ = \frac{{2 \times {\it{M}} \times {\it{N}}_A}}{{\frac{{{\mathrm{E}}_{{\mathrm{total}}}}}{{{\mathrm{E}}_{{\mathrm{photon}}}}}}} \times 100\% \\ = \frac{{2 \times {\it{M}} \times {\it{N}}_A}}{{\frac{{{\it{S}} \times {\it{P}} \times {\it{t}}}}{{{\it{h}} \times \frac{{\it{c}}}{{\it{\lambda }}}}}}} \times 100\% \\ = \frac{{2 \times {\it{M}} \times {\it{N}}_A \times {\it{h}} \times {\it{c}}}}{{{\it{S}} \times {\it{P}} \times {\it{t}} \times {\it{\lambda }}}} \times 100\% \end{array}$$
(1)

Where, Ne is the amount of generated electrons, Np is the incident photons, M is the amount of H2 molecules (mol), NA is Avogadro constant (6.022 × 1023 mol−1), h is the Planck constant (6.626 × 10-34 J·s), c is the speed of light (3 × 108 m s−1), S is the irradiation area (cm2), P is the intensity of irradiation light (W cm−2), t is the photoreaction time (s), λ is the wavelength of the monochromatic light (m).

## Data availability

All data supporting the findings of this study are available within the article, as well as the Supplementary Information file, or available from the corresponding authors on reasonable request.

## References

1. 1.

Perepichka, D. F. & Rosei, F. Extending polymer conjugation into the second dimension. Science 323, 216–217 (2009).

2. 2.

Ebrahimi, M. & Rosei, F. Organic analogues of graphene. Nature 542, 423 (2017).

3. 3.

Colson, J. W. & Dichtel, W. R. Rationally synthesized two-dimensional polymers. Nat. Chem. 5, 453 (2013).

4. 4.

Burroughes, J. H. et al. Light-emitting diodes based on conjugated polymers. Nature 347, 539 (1990).

5. 5.

Greenham, N. C., Moratti, S. C., Bradley, D. D. C., Friend, R. H. & Holmes, A. B. Efficient light-emitting diodes based on polymers with high electron affinities. Nature 365, 628 (1993).

6. 6.

Moreno, C. et al. Bottom-up synthesis of multifunctional nanoporous graphene. Science 360, 199–203 (2018).

7. 7.

Liu, W. et al. A two-dimensional conjugated aromatic polymer via C-C coupling reaction. Nat. Chem. 9, 563 (2017).

8. 8.

Zhuang, X. et al. A two-dimensional conjugated polymer framework with fully sp 2-bonded carbon skeleton. Polym. Chem. 7, 4176–4181 (2016).

9. 9.

Uribe-Romo, F. J. et al. A crystalline imine-linked 3-D porous covalent organic framework. J. Am. Chem. Soc. 131, 4570–4571 (2009).

10. 10.

Jin, E. et al. Two-dimensional sp2 carbon-conjugated covalent organic frameworks. Science 357, 673–676 (2017).

11. 11.

Zimmerman, H. E. & Ahramjian, L. Overlap control of carbanionoid reactions. II.1 The stereochemistry of the perkin reaction and related condensation reactions. J. Am. Chem. Soc. 81, 2086–2091 (1959).

12. 12.

Wang, X. et al. A metal-free polymeric photocatalyst for hydrogen production from water under visible light. Nat. Mater. 8, 76 (2008).

13. 13.

Sprick, R. S. et al. Tunable organic photocatalysts for visible-light-driven hydrogen evolution. J. Am. Chem. Soc. 137, 3265–3270 (2015).

14. 14.

Zhang, G., Lan, Z.-A. & Wang, X. Conjugated polymers: catalysts for photocatalytic hydrogen evolution. Angew. Chem. Int. Ed. 55, 2–18 (2016).

15. 15.

Yanagida, S., Kabumoto, A., Mizumoto, K., Pac, C. & Yoshino, K. Poly(para)phenylene-catalyzed photoreduction of water to hydrogen. J. Chem. Soc. Chem. Commun. 8, 474–475 (1985).

16. 16.

Shibata, T. et al. Novel visible-light-driven photocatalyst. Poly(p-phenylene)-catalyzed photoreductions of water, carbonyl compounds, and olefns. J. Phys. Chem. 94, 2068–2076 (1990).

17. 17.

Schwab, M. G. et al. Photocatalytic hydrogen evolution through fully conjugated poly(azomethine) networks. Chem. Commun. 46, 8932 (2010).

18. 18.

Yang, C. et al. Molecular engineering of conjugated polybenzothiadiazoles for enhanced hydrogen production by photosynthesis. Angew. Chem. Int. Ed. 55, 9202–9206 (2016).

19. 19.

Sprick, R. S. et al. Extended conjugated microporous polymers for photocatalytic hydrogen evolution from water. Chem. Commun. 52, 10008–10011 (2016).

20. 20.

Sprick, R. S. et al. Visible-light-driven hydrogen evolution using planarized conjugated polymer photocatalysts. Angew. Chem. Int. Ed. 55, 1792–1796 (2016).

21. 21.

Stegbauer, L., Schwinghammer, K. & Lotsch, B. V. A hydrazone-based covalent organic framework for photocatalytic hydrogen production. Chem. Sci. 5, 2789–2793 (2014).

22. 22.

Vyas, V. S. et al. A tunable azine covalent organic framework platform for visible light-induced hydrogen generation. Nat. Commun. 6, 8508 (2015).

23. 23.

Banerjee, T. et al. Single-site photocatalytic H2 evolution from covalent organic frameworks with molecular cobaloxime Co-catalysts. J. Am. Chem. Soc. 139, 16228–16234 (2017).

24. 24.

Pachfule, P. et al. Diacetylene functionalized covalent organic framework (COF) for photocatalytic hydrogen generation. J. Am. Chem. Soc. 140, 1423–1427 (2018).

25. 25.

Wei, P.-F. et al. Benzoxazole-linked ultrastable covalent organic frameworks for photocatalysis. J. Am. Chem. Soc. 140, 4623–4631 (2018).

26. 26.

Rao, M. R., Fang, Y., De Feyter, S. & Perepichka, D. F. Conjugated covalent organic frameworks via Michael addition-elimination. J. Am. Chem. Soc. 139, 2421–2427 (2017).

27. 27.

Friend, R. H. et al. Electroluminescence in conjugated polymers. Nature 397, 121 (1999).

28. 28.

Ng, S.-C. et al. Blue electroluminescence from a novel donor/acceptor polymer structure. Adv. Mater. 12, 1122–1125 (2000).

29. 29.

Attias, A.-J. et al. Columnar mesophase from a new disclike mesogen based on a 3,5-dicyano-2,4,6-tristyrylpyridine core. Chem. Mater. 14, 375–384 (2002).

30. 30.

Wang, H. et al. Synthesis and properties of new two-photon absorption chromophores containing 3,5-dicyano-2,4,6-tristyrylpyridine as the core. New J. Chem. 29, 792–797 (2005).

31. 31.

Peng, Y. et al. Ultrathin two-dimensional covalent organic framework nanosheets: preparation and application in highly sensitive and selective DNA detection. J. Am. Chem. Soc. 139, 8698–8704 (2017).

32. 32.

Dalapati, S., Jin, E., Addicoat, M., Heine, T. & Jiang, D. Highly emissive covalent organic frameworks. J. Am. Chem. Soc. 138, 5797–5800 (2016).

33. 33.

Zhang, T. et al. Copper-surface-mediated synthesis of acetylenic carbon-rich nanofibers for active metal-free photocathodes. Nat. Commun. 9, 1140 (2018).

34. 34.

Zhang, Y. & Antonietti, M. Photocurrent generation by polymeric carbon nitride solids: an initial step towards a novel photovoltaic system. Chem. Asian J. 5, 1307–1311 (2010).

35. 35.

Bian, J. et al. Efficiency enhancement of carbon nitride photoelectrochemical cells via tailored monomers design. Adv. Energy Mater. 6, 1600263 (2016).

36. 36.

Bi, S. et al. Substantial cyano-substituted fully sp 2-carbon-linked framework: metal-free approach and visible-light-driven hydrogen evolution. Adv. Funct. Mater. 27, 1703146 (2017).

37. 37.

Wang, X. et al. Sulfone-containing covalent organic frameworks for photocatalytic hydrogen evolution from water. Nat. Chem. 10, 1180–1189 (2018).

38. 38.

Wang, L. & Yamauchi, Y. Block copolymer mediated synthesis of dendritic platinum nanoparticles. J. Am. Chem. Soc. 131, 9152–9153 (2009).

39. 39.

Maeda, K. et al. Photocatalytic activities of graphitic carbon nitride powder for water reduction and oxidation under visible light. J. Phys. Chem. C. 113, 4940–4947 (2009).

40. 40.

Lan, Z.-A., Fang, Y., Zhang, Y. & Wang, X. Photocatalytic oxygen evolution from functional triazine-based polymers with tunable band structures. Angew. Chem. Int. Ed. 57, 470–474 (2018).

## Acknowledgements

We thank Shanghai Committee of Science and Technology for financial support within the project 16JC1400703. Furthermore, this work was financially supported by National Natural Science Foundation of China (21720102002, 21774072, 21574080, 21425309, 21761132002, 21861130353, and 21802022), the National Key Technologies R&D Program of China (2018YFA0209301) and the 111 Project (D16008). Open Project Program of the State Key Laboratory of Inorganic Synthesis and Preparative Chemistry (2019-01, Jilin University). S.B. acknowledges the Instrumental Analysis Center of SJTU for the NMR measurements. We thank Professor Feng Liu for his precious help.

## Author information

Authors

### Contributions

F.Z. and X.W. conceived and designed the experiments. S.B. performed the experiments and carried out the structural characterizations. C.Y. helped to analyze the results of photocatalytic experiments. W.Z. and D.W. helped with the structural characterization analysis. J.X. helped to synthesize the materials. L.L. and Y.H. helped to perform the high resolution TEM measurements. Q.L. performed the DFT calculations. S.B. and F.Z. wrote the manuscript with contributions from all the authors.

### Corresponding authors

Correspondence to Xinchen Wang or Fan Zhang.

## Ethics declarations

### Competing interests

The authors declare no competing interests.

Journal peer review information: Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work.

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## Rights and permissions

Reprints and Permissions

Bi, S., Yang, C., Zhang, W. et al. Two-dimensional semiconducting covalent organic frameworks via condensation at arylmethyl carbon atoms. Nat Commun 10, 2467 (2019). https://doi.org/10.1038/s41467-019-10504-6

• Accepted:

• Published:

• ### Catechol‐Coordinated Framework Film‐based Micro‐Supercapacitors with AC Line Filtering Performance

• Tianliang Yu
• , Youfu Wang
• , Kaiyue Jiang
• , Guangqun Zhai
• , Changchun Ke
• , Jichao Zhang
• , Jiantong Li
• , Diana Tranca
• , Emmanuel Kymakis
•  & Xiaodong Zhuang

Chemistry – A European Journal (2021)

• ### Recent developments in the photocatalytic applications of covalent organic frameworks: A review

• Junhua You
• , Yao Zhao
• , Lu Wang
•  & Wanting Bao

Journal of Cleaner Production (2021)

• ### Fabrication of Robust Covalent Organic Frameworks for Enhanced Visible-Light-Driven H2 Evolution

• Zhengfeng Zhao
• , Yunlong Zheng
• , Chun Wang
• , Sainan Zhang
• , Jie Song
• , Yafei Li
• , Shengqian Ma
• , Peng Cheng
• , Zhenjie Zhang
•  & Yao Chen

ACS Catalysis (2021)

• ### Covalent Organic Frameworks for Efficient Energy Electrocatalysis: Rational Design and Progress

• Hua Zhang
• , Minshen Zhu
• , Oliver G. Schmidt
• , Shuiliang Chen
•  & Kai Zhang

Advanced Energy and Sustainability Research (2021)

• ### Integrating active C3N4 moieties in hydrogen-bonded organic frameworks for efficient photocatalysis

• Tao Li
• , Bai-Tong Liu
• , Zhi-Bin Fang
• , Qi Yin
• , Rui Wang
•  & Tian-Fu Liu

Journal of Materials Chemistry A (2021)