ABSTRACT Chromoselective photocatalysis offers an intriguing opportunity to enable a specific reaction pathway out of a potentially possible multiplicity for a given substrate by using a

sensitizer that converts the energy of incident photon into the redox potential of the corresponding magnitude. Several sensitizers possessing different discrete redox potentials (high/low)

upon excitation with photons of specific wavelength (short/long) have been reported. Herein, we report design of molecular structures of two-dimensional amorphous covalent triazine-based

frameworks (CTFs) possessing intraband states close to the valence band with strong red edge effect (REE). REE enables generation of a continuum of excited sites characterized by their own

the intraband states to the conduction band of the CTF with 625 nm photons enables selective formation of C‒N cross-coupling products from arylhalides and pyrrolidine, while an undesirable

dehalogenation process is completely suppressed. SIMILAR CONTENT BEING VIEWED BY OTHERS COVALENT ORGANIC FRAMEWORKS WITH HIGH QUANTUM EFFICIENCY IN SACRIFICIAL PHOTOCATALYTIC HYDROGEN

EVOLUTION Article Open access 29 April 2022 CYCLOPALLADATION OF A COVALENT ORGANIC FRAMEWORK FOR NEAR-INFRARED-LIGHT-DRIVEN PHOTOCATALYTIC HYDROGEN PEROXIDE PRODUCTION Article Open access 10

February 2025 OVERCOMING SMALL-BANDGAP CHARGE RECOMBINATION IN VISIBLE AND NIR-LIGHT-DRIVEN HYDROGEN EVOLUTION BY ENGINEERING THE POLYMER PHOTOCATALYST STRUCTURE Article Open access 24

January 2024 INTRODUCTION Semiconductor photoredox catalysis opens up the possibility of realizing complex multistep chemical reactions under mild conditions1,2. In the past few years, a

booming development in conjugated polymers, starting from polymeric carbon nitride materials composed of heptazine units3,4,5, and covalent organic frameworks (COFs) with repeating molecular

units linked by strong covalent bonds, is witnessed6,7,8,9. Typical controlled synthetic routes toward COFs involve integrating building blocks with appropriate linkage, constructing a

π-conjugated system, and introducing electron donor–acceptor units into the framework10. This allows precise modulation of light absorption, band positions, as well as

intramolecular/intermolecular charge transport/separation behaviors. Vyas et al.11 reported, for example, a strategy for building an azine-linked COF platform with adjustable electronic and

structural properties for photocatalytic H2 production. Huang et al.12 demonstrated that the charge separation in a covalent triazine framework is promoted by altering electron

donor–acceptor moieties, which leads to the enhancement of activity in photoredox catalysis. Enhancement of COFs photocatalytic performance through inclusion of electron rich and electron

poor conjugates is a viable strategy, but application of such materials is limited to nonoxidizing environment and pH-values close to neutral13. One distinct advantage of COFs is their

regular nature, which enables the materials to serve as an ideal platform for structure engineering on a molecular level14. The feature of porosity and optoelectronic properties determined

by the molecular structure can thus be finely tuned to attain desired functionality15,16,17. The tunability, along with their extended π-conjugated framework, regular pore structure, and

high surface area, make COFs attractive alternatives to inorganic semiconductor photocatalysts18. Reduction and oxidation power of photocatalysts play a crucial role in enabling a specific

reaction. Ultimately, the chemical composition of a semiconductor determines the potential of conduction band (CB) and valence band (VB), which in turn define the driving force for

photoinduced electron transfer (PET) in photocatalytic applications19. Intriguing alternative to the abovementioned common strategies to adjust redox power of the photocatalytic system is

chromoselective photocatalysis, which has been proposed by Ghosh and König20. In their approach, illumination of a photocatalytic system composed of Rhodamine 6 G, a homogeneous sensitizer,

and electron donor with green or blue photons gives either moderately (‒1 V) or highly reductive (‒2.4 V vs SCE) species, respectively. In semiconductor carbon nitride photocatalysis,

Kroutil et al.21 and our group22 achieved various oxidation potentials of the excited state by selective excitation of n‒π* or π‒π* transitions. Overall, in chromoselective catalysis, one

specific reaction pathway (out of multiple possible) is enabled by the photon of specific wavelength23,24,25. Schematically a general principle behind the chromoselective photoredox

catalysis enabled by photons with shorter and longer wavelength is depicted in Fig. 1a. Despite plentiful reports on synthesis of crystalline COFs for photocatalytic H2 evolution15,26,27,

amorphous COFs (hereafter referred to as CTFs) offer a unique platform for designing materials with strong red edge effect (REE)28. REE conserves the information about the energy of incident

photons in the emission spectra, which is very different from majority of fluorophores in solution and crystalline semiconductors (Fig. 1b)29. REE was solely studied from a fundamental

perspective using, for example, graphene oxide30,31. To the best of our knowledge, using REE in the area of photocatalysis remains to this date rather elusive. Materials with REE break the

Kasha’s rule due to the existence of distribution of excited states on the timescale potentially sufficient for electron transfer (Fig. 1c). As a result, using incident photons with specific

wavelength would yield a continuum of excited sites with their own energies and redox potentials that may be used in chromoselective organic photoredox catalysis (Fig. 1b, c). Herein, we

present two-dimensional (2D) amorphous covalent triazine-based frameworks possessing strong REE. Nitrogen substitution in the CTFs peripheral aryl rings modulates porosity, optical gaps,

separation of charge carriers and density of intraband states. CTFs allow for enhanced visible-light-driven bromination of aromatic compounds using HBr and O2, even under low energetic 625 

nm photons. In dual Ni-photocatalysis, red light is essential to suppress completely the undesirable dehalogenation reaction and obtain selectively the product of C‒N cross coupling between

arylhalides and pyrrolidine. Structure of the synthesized CTFs and their activity in these two photocatalytic transformations is rationalized based on the spectroscopic study. High-energy

π‒π* and low-energy n‒π* transitions in the CTFs are responsible for the observed selectivity. RESULTS STRUCTURAL CHARACTERIZATION Two CTFs were synthesized by the polycondensation of

benzene-1,4-dicarboximidamide and different aldehydes (isophthalaldehyde and 2,6-pyridinedicarboxaldehyde) with procedures as described in ESI, denoted as PHT and PYT, respectively (Fig. 1d,

e). By varying the precursor aldehydes, CTFs with two distinct chemical structures are expected to be obtained, where the C-H moiety in the peripheral aryl rings in PHT is substituted with

one nitrogen atom in PYT (see the extended chemical structures in Supplementary Fig. 1). This designed structure endows PYT with different electron donor–acceptor domains compared with PHT

(marked in different color in Fig. 1d, e). The CTFs gave X-ray diffraction (XRD) patterns with rather broad peaks that indicate amorphous structures (Supplementary Fig. 2). As in most cases

of CTFs synthesis6,32,33, the disorder and amorphous phases formed during the condensation are likely responsible for the limited crystallinity. Both PHT and PYT show a main peak at around

25.1°, which could be ascribed to the π–π interlayer stacking motif corresponding to a distance of 0.354 nm. The additional peak at 18.0° in PYT may be assigned to the ultra-micropores (<

0.5 nm) in the CTF structure34. Additional information on the chemical structure of the CTFs was further gained through Fourier-transform infrared (FT-IR) spectroscopy, X-ray photoelectron

spectroscopy (XPS), and energy-dispersive X-ray analysis. The FT-IR spectrum of PHT shows two characteristic vibrations at 1514 and 1349 cm−1 (Fig. 2a), which can be attributed to the

aromatic C–N stretching and breathing modes in the triazine unit, respectively33. This result clearly supports formation of triazine linkages. The slight shift of the absorption band due to

C‒N stretching vibration in PYT indicates the change in C-N bond interaction due to the introduction of pyridinic nitrogen. In high-resolution XPS spectra (Fig. 2b–d), the C 1 _s_ peak at

the lower binding energy is assigned to the adventitious carbon (C–C) along with a contribution from the aromatic carbon (C=C) in the CTF framework, and is calibrated at 284.8 eV11. The

other three C 1 _s_ signals with binding energies at 285.9, 287.0, and 288.4 eV in PHT correspond to carbon atoms in C=N, C–O, and C–N chemical states, respectively35,36. The simultaneously

increased content of C=N and C–N groups in PYT implies formation of pyridine rings in the framework (see the integrated area ratios of the deconvoluted peaks in Supplementary Table 1). The

large N 1 _s_ peak centered at 398.7 eV is attributed to the pyridine nitrogen (C=N–C) from the triazine ring37, which is also present in the pyridine ring of PYT. The other peak at around

400.0 eV is assigned to the NHx groups from the edges of the frameworks38. The high-resolution O 1 s spectrum suggests that the surface of the CTFs contains C=O and C–O moieties, with

binding energies at 531.7 and 533.1 eV in PHT, respectively39, which have been incorporated as a result of partial terephtalamidine hydrolysis under alkaline conditions. An increased

proportion of the C–O chemical state was observed in PYT, indicating more functional groups with oxygen singly bound to carbon (e.g., C–OH, C–O–C, and C–OOH). Note that the C 1 _s_ and O 1 

_s_ peaks in PYT appear at the higher energy compared with PHT, indicating that the electron density of the C and O atoms has been shifted toward the electron-deficient pyridine ring. The

measured elemental composition of PYT and PHT agrees with the theoretical values (Supplementary Table 2). Incorporation of heteroatoms into the skeleton of COFs can significantly affect pore

structure40. The morphologies of the CTFs probed by scanning electron microscope (SEM) show that the materials consist of aggregated nanoparticles with the diameter >500 nm

(Supplementary Fig. 3). The transmission electron microscopy (TEM) confirmed amorphous structure of the synthesized CTFs, which is in agreement with the PXRD patterns or could be a result of

the CTFs instability under the electron beam (Supplementary Fig. 4)11,15. Compared with PHT, smaller and loosely packed aggregates were observed for PYT, implying a more porous structure.

The porosity of the CTFs was further investigated by N2 absorption–desorption measurements. The Brunauer–Emmett–Teller (BET) surface area (SBET) of PYT was found to be 104 m2 g−1, with a

pore volume of 0.36 cm3 g−1, higher than those of PHT (27 m2 g−1 and 0.06 cm3 g−1). Lower surfaces of the CTFs might stem form staking neighboring layers in AB fashion41,42,43. However, the

pore volume of PYT is comparable to that of the CTFs reported in the previous studies9,44,45. Pore size distribution was analyzed based on the quenched solid density functional theory model.

Pores with the similar diameter of around 2.0 nm were determined for both of the CTFs (Supplementary Fig. 5), in good agreement with the predicted width (1.9 nm) of the largest in-plane

pores created by the layers arranged in AA stacking (Supplementary Fig. 1). By virtue of the nanocrystalline nature of CTFs, a precise determination of the stacking type, that is, AA versus

AB, is impossible. Larger mesopores in the range of 5.0–16.0 nm observed for PYT are likely associated with the missing columnar π-walls – faults that occur during the synthesis. These

nanopores can offer an essential space and can potentially act as nanoreactors for catalytic reactions. Increased SBET and mesoporous structure of PYT are expected to be more favorable for

catalytic applications. Thermal gravimetric analysis measurements performed under a N2 atmosphere reveal that PHT and PYT are stable up to 300 °C without significant loss of mass (around

10%) (Supplementary Fig. 6). OPTICAL PROPERTIES AND EXCITED-STATE DYNAMICS The light absorption/emission of COFs is strongly dependent on the electron density distribution along the molecule

as well as the delocalization of the π-electrons throughout a π-conjugated backbone46. The rational introduction of substituents into a conjugated organic molecule, therefore, is a

practical means to manipulate the optical properties. As shown in Fig. 3a, the intrinsic absorption bands of the CTFs in the lower λ range are the result of π‒π* electron transitions in the

conjugated aromatic system47. The absorption bands located at higher λ are assigned to n‒π* electron transitions involving lone pairs of the nitrogen atoms of the triazine units48. PYT shows

a notably enhanced n‒π* absorption up to near IR (nIR) region compared with PHT, which is rationalized by the additional conjugation with the pyridinic nitrogen. This is consistent with the

color change, from yellowish to brownish (inset in Fig. 3a). A narrower optical band gap (_E_g) was observed for PYT (2.50 eV) in comparison to PHT (2.72 eV) (see corresponding Tauc plots

in Supplementary Fig. 7b). We attribute the shrinkage of the band gap to an increased planarity of the framework due to the nitrogen substitution, which leads to a higher level of

delocalization11,49,50. The VB levels in CTFs were determined from photoelectron spectra (Supplementary Fig. 8). The CB levels in CTFs were calculated by adding the _E_g values derived from

Tauc plots to the VB levels and are −3.51 and −2.93 eV for PHT and PYT, respectively (Fig. 3b). From Mott–Schottky plots (Supplementary Fig. 9), flat band potentials (_E_FB) were determined

as 0.17 V and 0.60 V below the CB edges in the CTFs. In the photoluminescence (PL) spectra of PHT and PYT, the position and intensity of emission maxima depend on the excitation wavelength,

which is known as red-edge effect (REE) (Fig. 3c, d)28. REE has been observed in numerous amorphous carbon-based rigid systems, such as graphene oxide and carbon quantum dots30,51. REE is

related to the existence of the excited-state distribution of fluorophores on their interaction energy with the environment. In other words, in such amorphous systems relaxation of the

excited state from higher to zero vibrational level occurs on the timescale longer than the emission of photons via fluorescence. The chemical structure of the synthesized CTFs affects the

REE. Thus, in the PL spectrum of PHT, the emission peak progressively shifts toward longer wavelengths upon increasing the excitation wavelength. The intensity of the PL signal maximizes at

527 nm upon excitation at 450 nm. In the PL spectrum of PYT, the fluorescence maximum also shifts toward longer wavelength along with the increase of λexc. But intensity progressively

increases along with λexc and maximizes at 715 nm upon excitation at 650 nm. These results agree well with the extended absorption edge of PYT, suggesting that the n‒π* absorption band

contributes to the fluorescence. Nanocrystaline CTF, PYTnc that was synthesized by polycondensation of the Schiff base of pyridine-2,6-dicarboxaldehyde and terephthalamidine dihydrochloride

(Supplementary Fig. 10), also demonstrates REE in the steady-state fluorescence spectra (Supplementary Note 1). Fluorescence intensity of PYT is significantly quenched compared with PHT upon

excitation at 450 nm, indicating a suppressed radiative charge carriers recombination (Supplementary Fig. 11)52. A possible explanation is that the larger electron-acceptor domains in PYT

comprising triazine and pyridine units with strong electron-withdrawing functionality render the photogenerated charge carrier separation more favorable. Similar phenomenon was observed in a

designed electron-transfer system composed of a cascade of electron acceptors52. To further probe the charge transfer process, time-resolved (TR)-PL spectra of the CTFs were recorded at

three λexc = 375, 470, and 640 nm, while monitoring the decay of photons with the wavelength in the range λem = 400–900 nm (Fig. 3e, Supplementary Fig. 12). The decay profiles were fitted

using a multiexponential decay model and \(\bar{{{{{{\rm{\tau }}}}}}}\) (amplitude average lifetime) is reported (see the calculation method in the Supporting Information)53. The fitting

parameters are summarized in Supplementary Table 3 and an example of the decay curve with instrument response function is shown in Supplementary Fig. 13. Taking into account low fluorescence

quantum yield (<0.1%), fluorescence spectroscopy does not characterize comprehensively the dynamics of the CTFs excited states. However, lifetimes (≤1 ns) extracted from fluorescence

decay curves correspond to the time of charge carriers separation. Analysis of data shown in Fig. 3e enables us to make the following conclusions related to the excited state dynamics in the

CTFs. REE strongly affects charge separation in the CTFs48,54. Thus, a faster separation of photogenerated charges is achieved upon excitation with more energetic photons such as λexc = 375

 nm compared with 470 nm or 640 nm (see also Supplementary Note 2 for additional explanation of TR-PL data). Upon excitation at 375 nm, PYT shows shorter \(\bar{{{{{{\rm{\tau }}}}}}}\), and

hence more efficient charge separation, compared with PHT. The situation is opposite upon excitation of CTFs at 640 nm, which leads to higher activity of PHT compared with PYT in the 

photocatalysis under red light (see below) (excited-state investigations by transient absorption spectroscopy revealed a complex deactivation behavior on a ps to ns timescale together with

excited-state dynamics exceeding lifetimes >50 µs (Supplementary Note 3). Possible rationales for such long-lived states are the formation of deep traps or free charge carriers population

long-lived surface states. The latter might be beneficial for application in photocatalysis that is based on energy transfer. Recently, the dependence of charge separation on excitation

wavelength in carbon nitrides possessing nitrogen vacancies close to VB has been investigated55. Furthermore, the CTFs were investigated by electron paramagnetic resonance (EPR) spectroscopy

(Fig. 3f). In the absence of light, both CTFs show the presence of stable organic species typical for polymeric networks with extended π-conjugation56. Specific concentration of free

radicals in PYT was determined to be 1.24 × 1017g−1, while in PHT it is 2.35 × 1016g−1. We relate the higher concentration of free radicals in PYT to higher nitrogen content that is due to

pyridinic moieties capable to stabilize free radicals in the π-conjugated framework57. Upon cooling to 90 K, the amplitude of the signal in EPR spectra of both CTFs increases, indicating a

paramagnetic ground state (Supplementary Fig. 14)58. In situ EPR spectra of CTFs recorded upon samples irradiation with visible light show gradual increase of signal amplitude. In case of

PHT, concentration of free radicals increases ca. 3.5 times and reaches saturation within first 5 min of irradiation, while PYT shows gradual accumulation of free radicals over the time and

after 40 min increases by ca. 2.2 times (Supplementary Fig. 15). This clear difference suggests an inhibited recombination and hence a prolonged lifetime of photoexcited charge carriers in

PYT. The results of spectroscopic study are summarized in Fig. 3g. Thus, excitation of the materials with photons, ≤450 nm for PHT (optical gap 2.72 eV) and ≤500 nm for PYT (optical gap 2.50

 eV), enables π‒π* transitions that lead to the formation of ‘hot’ electrons and holes followed by their migration in the π-conjugated structure of the materials to less energetic surface

states, which explains Stokes shift of ca. 50 nm (at λexc 365 nm). Excitation of CTFs with photons of lower energy, >450 nm for PHT and >500 nm for PYT, enables low energetic n‒π*

transitions. Amorphous (and nanocrystalline in case of PYTnc) structure of CTFs perturbs the electron transfer between nonbonding orbitals composed of nitrogen atoms in triazine (in case of

PHT) and in triazine/pyridine (in case of PYT). Therefore, for n‒π* transitions the holes remain localized at nitrogen atoms. The REE onset determined from steady-state emission spectra

correlates with the _E_g (defined by π‒π* transitions) of the CTFs – the narrower energy gap the more red-shifted the onset of REE (Supplementary Fig. 16). Analysis of absorption and

steady-state emission spectra also suggests that the density of the intraband states in the CTFs is different. Considering that the strongest fluorescence for PHT is registered upon

excitation at 450 nm (Fig. 3c), while the material does not absorb photons with λ > 800 nm (Fig. 3a), intraband states gradually evolve in the middle (from ca. ‒5.1 eV vs vacuum level) of

the PHT band gap and merge with the VB. In PYT, introduction of nitrogen atoms leads to the gradual evolution of intraband states starting from −4.1 eV as evidenced by absorption of the

material in nIR (Fig. 3a), while steady-state PL spectra (Fig. 3d) indicate localization of the states in the middle of the band at ca. ‒4.5 eV. Overall, incident photons generate a

population of oxidative states with the energy (versus vacuum level) in the range, relation (1): $${E}_{{{{{{\rm{IBSO}}}}}}}\,{ < }\,{{E}}_{{{{{{\rm{ox}}}}}}} \, < \,

\left({E}_{{{{{{\rm{CB}}}}}}}-\frac{1240}{{\lambda }_{{{{{{\rm{exc}}}}}}}}\right)$$ where _E_CB represents the energy of the CB minimum versus vacuum level, eV; _E_IBSO represents energy of

the intraband states onset versus vacuum level, eV; and λexc represents wavelength of incident photon, nm. The relation is valid as long as the energy of incident photons is sufficient for

excitation of either π‒π* or n‒π* transitions and defined by the Planck–Einstein relation (2): $$\frac{1240}{{\lambda }_{{{{{{\rm{exc}}}}}}}} \, > \;

({E}_{{{{{{\rm{CB}}}}}}}-{E}_{{{{{{\rm{IBSO}}}}}}})$$ Analysis of the spectroscopic data clearly indicates that despite the fact that chemical structures of PYT and PHT differ mainly in the 

nitrogen content, such differences have profound influence on materials properties – pore structure, optical gaps, separation of photogenerated charge carriers, and density of intraband

states. Such differences inspired us to explore these materials as photocatalysts in organic syntheses. APPLICATION IN OXIDATIVE HALOGENATION Here, we chose photocatalytic oxidative

bromination of aromatic compounds, since this method offers an alternative strategy for the synthesis of halogenated hydrocarbons – important intermediates in organic synthesis59,60. Anisole

was used as a model substrate, while HBr and O2 were used as the bromide source and electron-scavenger, respectively. In all experiments, 4-bromoanisole 2A was obtained as the only product

(Table 1). An 82% yield of the product was observed in the presence of PYT after 4 h of irradiation (entry 1), whereas PHT only gave a 7% yield (entry 2). Despite surface area of PYT is 4

times higher compared with PHT, such difference cannot account entirely for 12 times higher yield of anisole in the model reaction. PYTnc gave 4-bromoanisole as the only product (entry 3).

The yield was increased to 99% and 56% with PYT (entry 4) and PHT (entry 5), respectively, when the irradiation time was extended to 24 h. Low yields (2%–7%) of product were obtained without

photocatalyst (entry 6) and without light irradiation (entry 7). Even with lower amount of HBr (entry 8 and 9), we were able to observe a nearly 99% yield with PYT. Addition of variable

quantities of H2O2 into the reaction mixture gave 4-bromoanisole with lower yield, which may be explained by overoxidation of the product under such conditions (Supplementary Table 4). These

results obviously point at advantages of photocatalytic generation of H2O2 in situ. Despite quite strong acidic reaction conditions, no evident decay in the activity (entries 10–12) and no

significant structural difference (FT-IR, DRUV–vis, steady state, and TR PL spectra in Supplementary Figs. 17–20 and C, N, H elemental analysis in Supplementary Table 5) could be detected

after four cycles, manifesting the extraordinary stability of PYT. Results of control experiments to check for “leaching” suggest the lack of any organics derived from PYT, which are soluble

in acetonitrile and which could act as sensitizers in the oxidative bromination of anisole (Supplementary Table 6). The apparent quantum yield of the reaction was further measured and

determined to be 0.05% for PHT and 0.18% for PYT under blue light (455 nm), and 0.03% for PHT and 0.05% for PYT under green light (530 nm), respectively (Supplementary Fig. 21). Inspired by

the superior activity of PYT in the visible-light-driven bromination of anisole, we proceeded to examine the scope of the reaction. A series of substituted aromatics were selected as the

substrates (Fig. 4). In these experiments, a higher amount of substrate (0.6 mmol) was used to serve the purpose of examining the scalability of the reaction. 4-bromoanisole 2A was obtained

in excellent isolated yields (97%). The reaction was further scaled up to 6 mmol of 1A, while keeping the reaction time unchanged (48 h) via increasing the photon flux proportionally to the

amount of the substrate (Supplementary Fig. 23). In this case, the yield of 2A was 87%. 4-bromo-2-chloroanisole 2B was obtained as the only product using 2-chloroanisole 1B as the substrate,

with a good isolated yield of 80%, which is ascribed to the weakly electron withdrawing effect of the chlorine atom. The reaction with N,N-dimethylaniline 1C led to a mixture of 4-isomer

(2C) and 2,4-isomer (2Cʹ) in the presence of an electron-donating dimethylamino group. 1,3-dimethoxybenzene 1D was only converted into 2,4-isomer 2D due to the steric hindrance of the

methoxy groups, giving a high 95% isolated yield. 1,2,3-trimethoxybenzene 1E underwent bromination that gave 4-isomer 2E as the only product. In the case of 1,3,5-trimethoxybenzene 1 F,

almost the same amount of 4-isomer 2F and 2,4-brominated product 2Fʹ were obtained, giving a total yield of 81%. Using a mixture of H2O2 and HBr in dark, comparable yields of the brominated

products have been obtained for the same substrates (Supplementary Table 7), which implies similar mechanisms (see below). Both PYT and PHT enable photocatalytic bromination of

1,3,5-trimethoxybenzene, when KBr was used as bromide source without addition of acid (Supplementary Table 8)2. Using KBr as bromide source is beneficial for bromination of electron rich

acidophobic compounds, such as 3,4-ethylenedioxythiophene 1G, which gave 2,5-dibromo-3,4-ethylenedioxythiophene 2 G in 44% yield (Supplementary Table 9). In summary, bromination worked well

with the substrates having strong electron-donating groups (EDG), such as –OMe and –NMe2. In the presence of weak electron-withdrawing group (–Cl group), the reaction proceeded more slowly.

Besides, the steric effect of the substituents is also an important factor in the reaction selectivity of aromatic hydrocarbons. PYT was used as a model for unveiling the mechanism of the

photocatalytic bromination of anisole. As evidenced by EPR spectroscopy (Fig. 5a), addition of anisole to PYT led to the increase of the specific number of free radicals in the material,

from 2.69 × 1017g−1 to 2.81 × 1017g−1, upon irradiation. Therefore, in the studied reaction, anisole serves as a hole scavenger, which boosts the charge separation by electron transfer to

the CTF. It is well accepted that superoxide radical (O2•−) and singlet oxygen (1O2) are two main reactive oxygen species that play a decisive role in photocatalytic reactions61,62. Given

that the potential of PYT is located at −1.75 V vs. SCE (Supplementary Fig. 30), reduction of O2 to O2•− (O2/O2•−=−0.89 V vs. SCE)62 is thermodynamically favorable. To probe these two

species, EPR spectra of PYT in the dark and under light irradiation were monitored, with addition of 5,5-dimethyl-1-pyrroline N-oxide (DMPO) and 2,2,6,6-tetramethylpiperidine (TEMP) as O2•−

and 1O2 scavengers, respectively. As shown in Fig. 5b, O2•− trapping upon irradiation gave an EPR spectrum comprising asymmetric signals, which derive from the overlapped signals of DMPO-O2H

adduct (fit 1, hyperfine coupling constant 12.5 G) and radicals produced by PYT (fit 2)63. Since 1O2 was found to be hard to monitor, we decreased the PYT concentration in the PYT/TEMP

mixture (to eliminate the interference of radicals in PYT), and increased both the irradiation power and time to obtain a clear spectrum. Indeed, the appearance of the TEMPO signal confirmed

the formation of 1O2 (Fig. 5c). To determine the contribution of O2•− and 1O2 in the photocatalytic reaction, we carried out the reaction in the presence of benzoquinone (O2•− scavenger)

and sodium azide (NaN3) (1O2 scavenger). As a result, the product yield after 4 h was reduced from original 82% to 46% and 68%, respectively. This observation confirms the participation of

both oxygen species in the photocatalytic process, with O2•− playing a more dominant role than 1O2. To further elucidate the reaction dynamics of the excited PYT, the emission decay profiles

of PYT dispersion in MeCN/H2O (volume ratio 5:1) were recorded with λexc = 470 nm (photon energy close to that used in the photocatalytic experiments) and λem = 580 nm by picosecond TR-PL

spectroscopy. Figure 5d shows that the PL lifetime (amplitude average lifetime) of PYT dispersion is reduced from 74 to 56 ps under O2. The oxygen-induced quenching is ascribed to the

reaction of electrons with adsorbed O264. Effective quenching was also observed in the presence of anisole (Fig. 5e) and HBr (Fig. 5f), respectively. These data confirm the electron transfer

from both the reagents to the photoexcited PYT65, therefore implying a bromination via electrophilic and nucleophilic substitution pathways. The corresponding quenching constants (_k_q)

determined from the Stern-Volmer plot (inset in Fig. 4e and f) are 4.8 × 109 and 1.6 × 1010 M−1 s−1 in the case of anisole and HBr, respectively. Note that an intensity average lifetime was

used for the calculation of _k_q53 (see the calculation method in the Supporting Information). The _k_q values are close to the diffusion limit (~1010 M−1 s−1)66, and indicate a likely

faster electron transfer between excited PYT and HBr compared with anisole, presumably due to coupling with the transfer of a proton (proton-coupled electron transfer, PCET). A reaction

mechanism involving two possible (electrophilic and nucleophilic) pathways based on the above observations is shown in Supplementary Fig. 31 and discussed in Supplementary Note 4. Survey of

conditions typically employed in photocatalytic oxidative halogenation clearly indicates that the presence of protons in the reaction mixture is mandatory67. On the other hand, protonation

of semiconductors, such as carbon nitrides68 and ZnO69, shifts the VB potential to more positive values. In the context of oxidative bromination by CTFs, HBr not only provides protons for

reduction of O2 to HO2• or H2O2 via PCET (Supplementary Table 10), it protonates CTFs surfaces and as a result enables oxidation of more stable substrates (Supplementary Note 5). Indeed,

both PYT and PHT enable oxidative bromination of anisole in acidic environment, but not in neutral (Supplementary Table 11). CTFS IN DUAL NI-PHOTOCATALYTIC C‒N CROSS-COUPLING C‒C and

C‒heteroatom cross-coupling of arylhalides with secondary amines2,70, alcohols71,72,73, thiols74,75, trifluoroborates76, and alkenes77, respectively, mediated by a combination of cheap Ni

salts and heterogeneous photocatalysts free of platinum group metals offers a convenient and scalable approach for the synthesis of value-added organic compounds. We have chosen pyrrolidine

and 4-bromobenzonitrile 3A as reaction partners, NiBr2·glyme as precatalyst, and DABCO as base2. Upon illumination of the reaction mixture with 400 nm photons, the dehalogenation product 4A

was obtained selectively with 39% and 100% yield in case of PYT and PHT, respectively (Table 2, entries 1,2). Despite extended absorption, upon illumination with 465 nm photons PYT gave 4A

with 5% yield, but failed to produce 5A (entry 3). PHT and PYTnc gave 5A with 29% and 30%, respectively, but the main product was 4A (entry 4,5). Addition of phenantroline as a ligand to the

system containing PHT led to selective formation of 4A (entry 6). Without any Ni catalyst, PHT gave dehalogenation product with 44% yield (entry 7), while the reaction did not proceed

without any photocatalyst (entry 8). Similar to oxidative halogenation of anisole with PYT, a control experiment to check for “leaching” of organics from PHT that could sensitize the dual-Ni

photocatalytic C‒N cross-coupling was negative – neither 5A was formed nor any conversion of 3A took place. Under illumination with 525 nm photons for 120 h, PHT gave 5A with 76% yield

(entry 9), while PYT – only product 4A with 5% yield (entry 10). To suppress completely the formation of the dehalogenation product 4A, we switched to red (625 nm) photons. PHT gave 5A with

68% yield (entry 11). PYT did not convert any of the reagents (entry 12), while PYTnc gave 5A with 7% (entry 13). The yield of 5A was further improved to 89% without compromising the

selectivity by extending the reaction time under red light to 168 h (entry 14). Longer reaction times are quite common, when photons of longer wavelengths are employed in the dual

Ni-photocatalysis70. Optimizing, for example, the relative distance between the sensitizer and the transition metal site emerged as a viable strategy to improve the quantum efficiency of the

reaction. This has been demonstrated for an Ir-sensitizer and a Ni-catalyst integrated into a soft polymer78. After three rounds of use, PHT gave 5A with 80% yield (entry 15). A set of

techniques unambiguously confirmed that the chemical structure of PHT remained intact throughout the dual Ni-photocatalysis, which is explained by extremely mild conditions – low-energy

electromagnetic radiation (625 nm) and 25 °C (Supplementary Note 6). Overall, PHT enables selective coupling of electron-deficient arylhalides 3A-D with 18%–89% yields (Fig. 6), while

electron-rich substrates 3E-G failed to give the coupling products (Supplementary Fig. 32), which is one of the existing challenges in transition-metal catalyzed cross-coupling

reactions79,80. DISCUSSION The above results clearly speak for the efficacy of our designed strategy. Here, we discuss our understanding of how the molecular structural control of CTFs

translates into new properties and ultimately influences their photocatalytic activity. A photocatalytic reaction is initiated by photon absorption by a photocatalyst to generate

electron-hole pairs followed by their separation to drive redox reactions. As observed earlier11, the substitution of the C−H moiety with nitrogen atoms can induce an increase in planarity

of the platform, thus leading to a higher degree of conjugation and hence a redshift of the absorption. Indeed, nitrogen rich PYT possesses: i) a narrower optical gap of 2.50 eV compared

with 2.72 eV for PHT and ii) extended absorption in the nIR (Fig. 3a). Secondly, charge carriers can undergo rapid recombination before they can dissociate at the polymer/solution interface.

The construction of proper electron donor–acceptor systems, in our case, a system with larger acceptor domains, can be expected to promote the charge separation, thereby minimizing the

exciton loss before it decays back to the electronic ground state. As evidenced by TR-PL data (Fig. 3e, Supplementary Fig. 33) excited state dynamics in PHT, PYT, and PYTnc is different and

depends strongly on the excitation wavelength. As deduced from Stern-Volmer plots shown in Fig. 5e, f, electron transfer between PYT and anisole occurs at the rate of 4.8·109 M−1 s−1;

between PYT and HBr – 1.6·1010 M−1 s−1, which is limited by the diffusion of the reagents to the surface of the photocatalyst. Therefore, excitons that are separated with a rate smaller (or

time greater) than the rate of electron transfer do not enable redox reaction. Figure 7a shows the correlation between the yield of 4-bromoanisole and _the longest time_ (\({\bar{\tau

}}_{{\max }}\)) of photogenerated charge carriers separation upon excitation at 470 and 640 nm. Therefore, the faster separation of the photogenerated charge carriers the higher is the yield

of 4-bromoanisole under illumination with light of the corresponding wavelength. Despite the higher activity of photocatalysts (higher yield of a product) in certain reactions is typically

explained by improved absorption of photons of certain wavelength, our results indicate that absorption itself may not be the only factor. Despite PHT has significantly lower absorption at

620–625 nm compared with PYT (Fig. 3a), it produces 4-bromoanisole with a higher rate, that is, 27% yield versus 12% (Table 1). Potentials of the occupied states, VB or intraband states

laying close to the VB, define the driving force for electron transfer from the substrate. Therefore, selective excitation of either n‒π* or π‒π* transitions in the conjugated polymers, such

as carbon nitrides, can alter the selectivity of a chemical reaction21,22. Analyses of TCSPC data and steady-state emission spectra in combination with the results of CTFs photocatalytic

activity in oxidative bromination and dual Ni-photocatalytic C-N cross-coupling suggest that the concept is applicable for our systems as well. Thus, excitation of CTFs with shorted

wavelength, such as λ ≤ 450 nm for PHT, λ ≤ 500 nm for PYT and λ ≤ 650 nm for PYTnc as deduced from Tauc plots (Supplementary Fig. 7b), gives the excited state with π‒π* character. On the

other hand, excitation of CTFs with longer wavelength, such as λ > 450 nm for PHT, λ > 500 nm for PYT and λ > 650 nm for PYTnc, give excited state with n‒π* character. These excited

states possess oxidation power that is defined by the band structure and the energy of incident photons (Relation 1). Whenever a reaction occurs at the interface (e.g., heterogeneous

catalysis), porosity may play a critical role. Some studies did not show a direct correlation between the degree of porosity and activity16, while others reported a marked effect81. In the

context of oxidative bromination, an increase in surface area and pore volume in PYT may favor photocatalytic reaction by providing a larger number of basic sites for hosting protons, which

facilitate oxidation of the substrates via multisite PCET – protons are supplied by HBr, while electrons by the substrate. Larger surface area and higher number of basic sites in PYT

compared with PHT, facilitates trapping of electrons under ambient conditions as deduced from EPR spectra (Fig. 3f). In the photocatalytic cycle, compensation of the excessive negative

charge on the surface of the photocatalyst by coupling it with H+ stabilizes the system by ΔGst, which for PYT is as high as 1.06 eV when taken into account the oxidation potentials of the

substrates from Fig. 3 and VB potential of CTFs (Fig. 7b, Supplementary Fig. 30). As a result, energy of photons at the PET step is directed toward the oxidation of thermodynamically more

stable substrates. However, porosity is neither the only nor the primary factor that determines the CTF performance. If this would be the case, due to, for example, improved accessibility of

the substrates, PYT, which features a four times larger surface area and six times larger cumulative pore volume than PHT, would be more active in all of the photocatalytic experiments.

Moreover, PYT would be more active than PHT regardless of the photon wavelength, which was selected for the photocatalytic experiments. This is, however, not the case (Supplementary Fig. 

34). In fact, the CTF activity correlates with its surface area and porosity only in the oxidative halogenation of anisole under 468 nm illumination. But, PHT, which has a lower surface area

as well as lower porosity, is more active in oxidative halogenation under 625 nm illumination and dual Ni-photocatalytic C‒N coupling under 465, 525, and 625 nm illumination. In the case of

C‒N cross-coupling, analysis shown in Fig. 6a did not give a trend likely because the reaction is mediated by Ni-catalyst, while CTF in combination with light facilitate destabilization of

Ni-intermediates postulated in the catalytic cycle either via electron or energy transfer82,83. In particular, reductive elimination from Ni(II)-intermediate is endothermic84,85, but becomes

exothermic if Ni is in oxidation state III (Supplementary Fig. 35). Several conclusions related to the band structure of CTFs and their activity could be made taking into account mechanism

of dual Ni-photocatalytic C‒N cross-coupling suggested in the literature, which is based on one-electron transfer86. Overall, relative alignment of band edges in PHT, which upon photon

absorption gives the excited state that is more oxidative and less reductive compared with PYT, is beneficial to generate tentative Ni(I) and Ni(III) species (Supplementary Note 7). On the

other hand, as inferred from transient absorption spectroscopy, the lifetime of the excited states exceeds >50 μs for all CTFs studied herein (Supplementary Note 3). Therefore, energy

transfer pathway cannot be excluded87,88. Earlier, Pieber et al. have shown that wavelength of incident photons is one of the leverages to decrease the rate of reductive elimination in the

catalytic cycle and therefore mitigate deactivation of Ni-catalyst70. Our results, presented in Table 2, indicate that by decreasing the energy of incident photons, we effectively suppress

undesirable dehalogenation process and improve selectivity toward the product of C‒N cross-coupling. We believe that the photocatalytic reaction is a complex process where multiple factors

are at play. Our study suggests that not only the structure of amorphous carbon-based materials, but also the wavelength of incident photons strongly affect charge separation and redox

properties. As such tuning the chemical structure of the material and selection of the light source with appropriate wavelength can be used to tune performance of the material in the

photocatalytic reaction. In conclusions, two amorphous CTFs with extended light absorption in the visible range were prepared, instrumentalizing the gap between the conduction and valence

band edges with the intraband states that originate from nitrogen lone pairs embedded into the framework. In PHT, which possesses p-phenylene linkers, the intraband states are located in the

vicinity of the VB, while in PYT possessing pyridine-2,6-diyl linkers they are in the middle of the band gap. The intraband states are weakly emissive and photocatalytically active. Two

reactions, oxidative bromination of electron-rich aromatic compounds and dual Ni-photocatalytic C‒N cross-coupling, are mediated by the prepared CTFs. In oxidative bromination, the intraband

states allow the reaction to proceed under illumination with 625 nm photons. The yield of the products correlates with the time of photogenerated charge carriers separation. The role of

acidic medium in oxidative halogenation has been explained by a stabilizing effect of H+ on the reductively quenched photocatalyst when coupled with the transfer of electron. Due to the 

relative alignment of band edges in PHT that upon excitation gives more oxidative and less reductive excited state compared with PYT, CTF mediates dual Ni-photocatalytic C‒N cross-coupling

of electron-deficient arylhalides and pyrrolidine. Using 625 nm photons, we suppressed completely the undesirable dehalogenation process and maximized the yield of the cross-coupling

product. The mechanism of photocatalytic reactions mediated by CTFs is complex, but due to several features, such as relative alignment of band edges, the presence of intraband states, and

an ensemble of basic pyridinic moieties, they offer several modes for substrate activation, which in homogeneous photocatalysis are typically achieved by combining several molecular

mediators. We envision that progress in controlled synthetic approaches combined with careful structural and photophysical characterizations are promising avenues to the establishment of

precise structure-property relationships in nanomaterials, which will undoubtedly increase reaction control and thereby pave the way to large-scale applications of polymer-based

photocatalysts. METHODS SYNTHESIS OF PHT AND PYT The terephthalamidine dihydrochloride precursors were synthesized based on reported procedure89. Terephthalamidine dihydrochloride (1177.5 

mg, 5 mmol) and either 2,6-pyridinedicarboxaldehyde (337.8 mg, 2.5 mmol) (for PYT) or isophthalaldehyde (335.3 mg, 2.5 mmol) for (PHT), and cesium carbonate (3580 mg, 11 mmol) were added to

a solution of DMSO (25.0 mL) and water (1 mL) in 50 mL round-bottom flask. The mixture was heated at 100 °C for 24 h, then up to 150 °C for 36 h. After cool down to room temperature the

resulting precipitate was washed with diluted HCl (3 × 20 mL) to remove the salt and residual cesium carbonate, and washed with water (3 × 30 mL), acetone (3 × 10 mL), and THF (3 × 10 mL),

before drying at 80 °C under vacuum for 12 h. Finally, PYT and PHT were obtained as yellow powders. PHOTOCATALYTIC OXIDATIVE BROMINATION OF ANISOLE (SCREENING REACTIONS) A glass vial (4 mL)

was charged with a mixture of anisole (2.2 μL, 20 μmol), HBr (0.1 mL, 48 wt.%), photocatalyst (4 mg), and acetonitrile (0.5 mL). Magnetic stir bar was placed in the vial. The vial was purged

with O2 for 30 s and closed with cap. The reaction mixture was vigrously stirred in the photoreactor with blue LED module 1 (14 mW cm−2) for 24 h. After the reaction, CDCl3 (0.7 mL) and

water (0.1 mL) were added to the reaction mixture. The organic layer was separated and measured with 1H NMR using 1,4-dinitrobenzene as internal standard. The catalyst in aqueous phase was

separated by centrifugation, washed with acetonitrile for three times, and dried for recycle use. DATA AVAILABILITY The datasets generated during and/or analyzed during the current study are

available from the corresponding author on reasonable request. REFERENCES * Savateev, A. & Antonietti, M. Heterogeneous organocatalysis for photoredox chemistry. _ACS Catal._ 8,

9790–9808 (2018). Article  CAS  Google Scholar  * Ghosh, I. et al. Organic semiconductor photocatalyst can bifunctionalize arenes and heteroarenes. _Science_ 365, 360–366 (2019). Article 

ADS  CAS  PubMed  Google Scholar  * Savateev, A., Ghosh, I., Konig, B. & Antonietti, M. Photoredox catalytic organic transformations using heterogeneous carbon nitrides. _Angew. Chem.

Int Ed._ 57, 15936–15947 (2018). Article  CAS  Google Scholar  * Vilela, F., Zhang, K. & Antonietti, M. Conjugated porous polymers for energy applications. _Energ. Environ. Sci._ 5,

7819–7832 (2012). Article  CAS  Google Scholar  * Krishnaraj, C., Jena, H. S., Leus, K. & Van der Voort, P. Covalent triazine frameworks - a sustainable perspective. _Green. Chem._ 22,

1038–1071 (2020). Article  Google Scholar  * Beuerle, F. & Gole, B. Covalent organic frameworks and cage compounds: design and applications of polymeric and discrete organic scaffolds.

_Angew. Chem. Int. Ed. Engl._ 57, 4850–4878 (2018). Article  CAS  PubMed  Google Scholar  * El-Kaderi, H. M. et al. Designed synthesis of 3D covalent organic frameworks. _Science_ 316,

268–272 (2007). Article  ADS  CAS  PubMed  Google Scholar  * Ding, S. Y. & Wang, W. Covalent organic frameworks (COFs): from design to applications. _Chem. Soc. Rev._ 42, 548–568 (2013).

Article  CAS  PubMed  Google Scholar  * Wang, Z. J., Ghasimi, S., Landfester, K. & Zhang, K. A. Molecular structural design of conjugated microporous poly(benzooxadiazole) networks for

enhanced photocatalytic activity with visible light. _Adv. Mater._ 27, 6265–6270 (2015). Article  CAS  PubMed  Google Scholar  * Yang, Q., Luo, M., Liu, K., Cao, H. & Yan, H. Covalent

organic frameworks for photocatalytic applications. _Appl. Catal. B-Environ._ 276, 119174 (2020). Article  CAS  Google Scholar  * Vyas, V. S. et al. A tunable azine covalent organic

framework platform for visible light-induced hydrogen generation. _Nat. Commun._ 6, 8508 (2015). Article  ADS  CAS  PubMed  Google Scholar  * Huang, W. et al. Asymmetric covalent triazine

framework for enhanced visible-light photoredox catalysis via energy transfer cascade. _Angew. Chem. Int. Ed. Engl._ 57, 8316–8320 (2018). Article  CAS  PubMed  Google Scholar  * Haase, F.

& Lotsch, B. V. Solving the COF trilemma: towards crystalline, stable and functional covalent organic frameworks. _Chem. Soc. Rev._ 49, 8469–8500 (2020). Article  CAS  PubMed  Google

Scholar  * Wang, L. et al. Structural design principle of small-molecule organic semiconductors for metal-free, visible-light-promoted photocatalysis. _Angew. Chem. Int. Ed._ 55, 9783–9787

(2016). Article  CAS  Google Scholar  * Bi, S. et al. Two-dimensional semiconducting covalent organic frameworks via condensation at arylmethyl carbon atoms. _Nat. Comm._ 10, 2467 (2019).

Article  ADS  Google Scholar  * Sprick, R. S. et al. Tunable organic photocatalysts for visible-light-driven hydrogen evolution. _J. Am. Chem. Soc._ 137, 3265–3270 (2015). Article  CAS 

PubMed  Google Scholar  * Walczak, R. et al. Template- and metal-free synthesis of nitrogen-rich nanoporous “noble” carbon materials by direct pyrolysis of a preorganized hexaazatriphenylene

precursor. _Angew. Chem. Int. Ed._ 57, 10765–10770 (2018). Article  CAS  Google Scholar  * Wei, P. F. et al. Benzoxazole-linked ultrastable covalent organic frameworks for photocatalysis.

_J. Am. Chem. Soc._ 140, 4623–4631 (2018). Article  CAS  PubMed  Google Scholar  * 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. Engl._ 57, 470–474 (2018). Article  CAS  PubMed  Google Scholar  * Ghosh, I. & Konig, B. Chromoselective

photocatalysis: controlled bond activation through light-color regulation of redox potentials. _Angew. Chem. Int. Ed._ 55, 7676–7679 (2016). Article  CAS  Google Scholar  * Kroutil, W. et

al. Chromoselective photocatalysis enables stereocomplementary biocatalytic pathways. _Angew. Chem. Int. Ed_. 60, 6965–6969 (2021). Article  CAS  Google Scholar  * Markushyna, Y. et al.

Chromoselective synthesis of sulfonyl chlorides and sulfonamides with potassium poly(heptazine imide) photocatalyst. _Angew. Chem. Int. Ed._ 60, 20543–20550 (2021). Article  CAS  Google

Scholar  * Protti, S., Ravelli, D. & Fagnoni, M. Wavelength dependence and wavelength selectivity in photochemical reactions. _Photochemical Photobiological Sci._ 18, 2094–2101 (2019).

Article  CAS  Google Scholar  * Dai, Y. et al. Light-tuned selective photosynthesis of azo- and azoxy-aromatics using graphitic C3N4. _Nat. Commun._ 9, 60 (2018). Article  ADS  CAS  PubMed 

PubMed Central  Google Scholar  * Martinez-Gualda, A. M. et al. Chromoselective access to Z- or E- allylated amines and heterocycles by a photocatalytic allylation reaction. _Nat. Commun._

10, 2634 (2019). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Pütz, A. M. et al. Total scattering reveals the hidden stacking disorder in a 2D covalent organic framework.

_Chem. Sci._ 11, 12647–12654 (2020). Article  PubMed  PubMed Central  Google Scholar  * Xu, J. et al. Vinylene‐linked covalent organic frameworks (COFs) with symmetry‐tuned polarity and

photocatalytic activity. _Angew. Chem. Int Ed._ 132, 24053–24061 (2020). Article  ADS  Google Scholar  * Demchenko, A. P. The red-edge effects: 30 years of exploration. _Luminescence_ 17,

19–42 (2002). Article  CAS  PubMed  Google Scholar  * Demchenko, A. P., Tomin, V. I. & Chou, P. T. Breaking the Kasha rule for more efficient photochemistry. _Chem. Rev._ 117,

13353–13381 (2017). Article  CAS  PubMed  Google Scholar  * Cushing, S. K. et al. Excitation wavelength dependent fluorescence of graphene oxide controlled by strain. _Nanoscale_ 9,

2240–2245 (2017). Article  CAS  PubMed  Google Scholar  * Cushing, S. K., Li, M., Huang, F. & Wu, N. Origin of strong excitation wavelength dependent fluorescence of graphene oxide. _ACS

Nano_ 8, 1002–1013 (2014). Article  CAS  PubMed  Google Scholar  * Nguyen, V. & Grunwald, M. Microscopic origins of poor crystallinity in the synthesis of covalent organic framework

COF-5. _J. Am. Chem. Soc._ 140, 3306–3311 (2018). Article  CAS  PubMed  Google Scholar  * Huang, W. et al. Visible-light-promoted selective oxidation of alcohols using a covalent triazine

framework. _ACS Catal._ 7, 5438–5442 (2017). Article  CAS  Google Scholar  * Zhao, Y. F., Yao, K. X., Teng, B. Y., Zhang, T. & Han, Y. A perfluorinated covalent triazine-based framework

for highly selective and water-tolerant CO2 capture. _Energy Environ. Sci._ 6, 3684–3692 (2013). Article  CAS  Google Scholar  * Shen, Y. et al. Defect-abundant covalent triazine frameworks

as sunlight-driven self-cleaning adsorbents for volatile aromatic pollutants in water. _Environ. Sci. Technol._ 53, 9091–9101 (2019). Article  ADS  CAS  PubMed  Google Scholar  * Kurpil, B.,

Otte, K., Antonietti, M. & Savateev, A. Photooxidation of N-acylhydrazones to 1,3,4-oxadiazoles catalyzed by heterogeneous visible-light-active carbon nitride semiconductor. _Appl.

Catal. B: Environ._ 228, 97–102 (2018). Article  CAS  Google Scholar  * Zhu, X. et al. A superacid-catalyzed synthesis of porous membranes based on triazine frameworks for CO2 separation.

_J. Am. Chem. Soc._ 134, 10478–10484 (2012). Article  CAS  PubMed  Google Scholar  * Dontsova, D. et al. Triazoles: a new class of precursors for the synthesis of negatively charged carbon

nitride derivatives. _Chem. Mater._ 27, 5170–5179 (2015). Article  CAS  Google Scholar  * Lu, Z. Y. et al. High-efficiency oxygen reduction to hydrogen peroxide catalysed by oxidized carbon

materials. _Nat. Catal._ 1, 156–162 (2018). Article  CAS  Google Scholar  * Guo, L. et al. Engineering heteroatoms with atomic precision in donor–acceptor covalent triazine frameworks to

boost photocatalytic hydrogen production. _J. Mater. Chem. A_ 6, 19775–19781 (2018). Article  CAS  Google Scholar  * Huang, L. et al. Solvothermal synthesis of covalent triazine framework

and its application in photodegradation of organic dyes. _Mater. Today Chem._ 20, 100475 (2021). Article  CAS  Google Scholar  * Yang, Z. et al. Transformation strategy for highly

crystalline covalent triazine frameworks: from staggered AB to eclipsed AA stacking. _J. Am. Chem. Soc._ 142, 6856–6860 (2020). Article  CAS  PubMed  Google Scholar  * Ren, S. et al. Porous,

fluorescent, covalent triazine-based frameworks via room-temperature and microwave-assisted synthesis. _Adv. Mater._ 24, 2357–2361 (2012). Article  ADS  CAS  PubMed  Google Scholar  * Kuhn,

P., Antonietti, M. & Thomas, A. Porous, covalent triazine-based frameworks prepared by ionothermal synthesis. _Angew. Chem. Int Ed._ 47, 3450–3453 (2008). Article  CAS  Google Scholar 

* Zhang, K., Kopetzki, D., Seeberger, P. H., Antonietti, M. & Vilela, F. Surface area control and photocatalytic activity of conjugated microporous poly(benzothiadiazole) networks.

_Angew. Chem. Int. Ed._ 125, 1472–1476 (2013). Article  ADS  Google Scholar  * Bronstein, H., Nielsen, C. B., Schroeder, B. C. & McCulloch, I. The role of chemical design in the

performance of organic semiconductors. _Nat. Rev. Chem._ 4, 66–77 (2020). Article  CAS  Google Scholar  * Zhang, J. et al. Synthesis of a carbon nitride structure for visible-light catalysis

by copolymerization. _Angew. Chem. Int. Ed._ 49, 441–444 (2010). Article  CAS  Google Scholar  * Zhang, G., Savateev, A., Zhao, Y., Li, L. & Antonietti, M. Advancing the n → π* electron

transition of carbon nitride nanotubes for H2 photosynthesis. _J. Mater. Chem. A_ 5, 12723–12728 (2017). Article  CAS  Google Scholar  * Cho, H.-H. et al. Effect of incorporated nitrogens

on the planarity and photovoltaic performance of donor–acceptor copolymers. _Macromolecules_ 45, 6415–6423 (2012). Article  ADS  CAS  Google Scholar  * Conboy, G. et al. To bend or not to

bend – are heteroatom interactions within conjugated molecules effective in dictating conformation and planarity? _Mater. Horiz._ 3, 333–339 (2016). Article  CAS  Google Scholar  * LeCroy,

G. E. et al. Characteristic excitation wavelength dependence of fluorescence emissions in carbon “quantum” dots. _J. Phys. Chem. C._ 121, 28180–28186 (2017). Article  CAS  Google Scholar  *

Guo, L., Niu, Y., Razzaque, S., Tan, B. & Jin, S. Design of D–A1–A2 covalent triazine frameworks via copolymerization for photocatalytic hydrogen evolution. _ACS Catal._ 9, 9438–9445

(2019). Article  CAS  Google Scholar  * Sillen, A. & Engelborghs, Y. The correct use of “average” fluorescence parameters. _Photochem. Photobio._ 67, 475–486 (1998). CAS  Google Scholar

  * Zhang, S. et al. MoS2 quantum dot growth induced by S vacancies in a ZnIn2S4 monolayer: atomic-level heterostructure for photocatalytic hydrogen production. _ACS Nano_ 12, 751–758

(2018). Article  CAS  PubMed  Google Scholar  * Xue, J., Fujitsuka, M. & Majima, T. The role of nitrogen defects in graphitic carbon nitride for visible-light-driven hydrogen evolution.

_Phys. Chem. Chem. Phys._ 21, 2318–2324 (2019). Article  CAS  PubMed  Google Scholar  * Fischer, S., Schmidt, J., Strauch, P. & Thomas, A. An anionic microporous polymer network prepared

by the polymerization of weakly coordinating anions. _Angew. Chem. Int Ed._ 52, 12174–12178 (2013). Article  CAS  Google Scholar  * Bildirir, H., Paraknowitsch, J. P. & Thomas, A. A

tetrathiafulvalene (TTF)-conjugated microporous polymer network. _Chemistry_ 20, 9543–9548 (2014). Article  CAS  PubMed  Google Scholar  * London, A. E. et al. A high-spin ground-state

donor-acceptor conjugated polymer. _Sci. Adv._ 5, eaav2336 (2019). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Li, R. et al. Photocatalytic selective bromination of

electron-rich aromatic compounds using microporous organic polymers with visible light. _ACS Catal._ 6, 1113–1121 (2016). Article  ADS  CAS  Google Scholar  * Markushyna, Y. et al.

Halogenation of aromatic hydrocarbons by halide anion oxidation with poly(heptazine imide) photocatalyst. _Appl Catal. B-Environ._ 248, 211–217 (2019). Article  CAS  Google Scholar  *

Nosaka, Y. & Nosaka, A. Y. Generation and detection of reactive oxygen species in photocatalysis. _Chem. Rev._ 117, 11302–11336 (2017). Article  CAS  PubMed  Google Scholar  * Savateev,

A. et al. Potassium poly(heptazine imide): transition metal-free solid-state triplet sensitizer in cascade energy transfer and [3+2]-cycloadditions. _Angew. Chem. Int Ed._ 59, 15061–15068

(2020). Article  CAS  Google Scholar  * Reszka, K., Bilski, P. & Chignell, C. F. EPR spectra of DMPO spin adducts of superoxide and hydroxyl radicals in pyridine. _Free Radic. Res

Commun._ 17, 377–385 (1992). Article  CAS  PubMed  Google Scholar  * Yamakata, A., Ishibashi, T. & Onishi, H. Water- and oxygen-induced decay kinetics of photogenerated electrons in TiO2

and Pt/TiO2: A time-resolved infrared absorption study. _J. Phys. Chem. B_ 105, 7258–7262 (2001). Article  CAS  Google Scholar  * Choi, J. R., Tachikawa, T., Fujitsuka, M. & Majima, T.

Europium-based metal-organic framework as a photocatalyst for the one-electron oxidation of organic compounds. _Langmuir_ 26, 10437–10443 (2010). Article  CAS  PubMed  Google Scholar  *

Martinez-Haya, R., Miranda, M. A. & Marin, M. L. Metal-free photocatalytic reductive dehalogenation using visible-light: a time-resolved mechanistic study. _Eur. J. Org. Chem._ 2017,

2164–2169 (2017). Article  CAS  Google Scholar  * Podgoršek, A., Zupan, M. & Iskra, J. Oxidative halogenation with “green” oxidants: oxygen and hydrogen peroxide. _Angew. Chem. Int. Ed._

48, 8424–8450 (2009). Article  CAS  Google Scholar  * Ye, C. et al. Enhanced driving force and charge separation efficiency of protonated g-C3N4 for photocatalytic O2 evolution. _ACSCatal._

5, 6973–6979 (2015). CAS  Google Scholar  * Valdez, C. N., Schimpf, A. M., Gamelin, D. R. & Mayer, J. M. Proton-controlled reduction of ZnO nanocrystals: effects of molecular

reductants, cations, and thermodynamic limitations. _J. Am. Chem. Soc._ 138, 1377–1385 (2016). Article  CAS  PubMed  Google Scholar  * Gisbertz, S., Reischauer, S. & Pieber, B.

Overcoming limitations in dual photoredox/nickel-catalysed C–N cross-couplings due to catalyst deactivation. _Nat. Catal._ 3, 611–620 (2020). Article  CAS  Google Scholar  * Malik, J. A.,

Madani, A., Pieber, B. & Seeberger, P. H. Evidence for photocatalyst involvement in oxidative additions of nickel-catalyzed carboxylate o-arylations. _J. Am. Chem. Soc._ 142, 11042–11049

(2020). Article  CAS  PubMed  PubMed Central  Google Scholar  * Pieber, B. et al. Semi-heterogeneous dual nickel/photocatalysis using carbon nitrides: esterification of carboxylic acids

with aryl halides. _Angew. Chem. Int. Ed._ 58, 9575–9580 (2019). Article  CAS  Google Scholar  * Zhao, X. et al. Nickel-coordinated carbon nitride as a metallaphotoredox platform for the

cross-coupling of aryl halides with alcohols. _ACS Catal._ 10, 15178–15185 (2020). Article  CAS  Google Scholar  * Cavedon, C., Madani, A., Seeberger, P. H. & Pieber, B.

Semiheterogeneous dual nickel/photocatalytic (thio)etherification using carbon nitrides. _Org. Lett._ 21, 5331–5334 (2019). Article  CAS  PubMed  PubMed Central  Google Scholar  * Chen, H.

et al. A visible-light-harvesting covalent organic framework bearing single nickel sites as a highly efficient sulfur–carbon cross-coupling dual catalyst. _Angew. Chem. Int. Ed._ 60,

10820–10827 (2021). Article  ADS  CAS  Google Scholar  * Khamrai, J., Ghosh, I., Savateev, A., Antonietti, M. & König, B. Photo-Ni-Dual-Catalytic C(sp2)–C(sp3) cross-coupling reactions

with mesoporous graphitic carbon nitride as a heterogeneous organic semiconductor photocatalyst. _ACS Catal._ 10, 3526–3532 (2020). Article  CAS  Google Scholar  * Khamrai, J., Das, S.,

Savateev, A., Antonietti, M. & König, B. Mizoroki–Heck type reactions and synthesis of 1,4-dicarbonyl compounds by heterogeneous organic semiconductor photocatalysis. _Green. Chem._ 23,

2017–2024 (2021). Article  CAS  Google Scholar  * Pan, Y. et al. Boosting photocatalytic activities for organic transformations through merging photocatalyst and transition-metal catalyst in

flexible polymers. _ACS Catal._ 10, 11758–11767 (2020). Article  CAS  Google Scholar  * Qi, Z.-H. & Ma, J. Dual role of a photocatalyst: generation of Ni(0) catalyst and promotion of

catalytic C–N bond formation. _ACS Catal._ 8, 1456–1463 (2018). Article  CAS  Google Scholar  * Kawamata, Y. et al. Electrochemically driven, ni-catalyzed aryl amination: scope, mechanism,

and applications. _J. Am. Chem. Soc._ 141, 6392–6402 (2019). Article  CAS  PubMed  PubMed Central  Google Scholar  * Yang, C. et al. Molecular engineering of conjugated polybenzothiadiazoles

for enhanced hydrogen production by photosynthesis. _Angew. Chem. Int Ed._ 55, 9202–9206 (2016). Article  CAS  Google Scholar  * Zhu, C., Yue, H., Jia, J. & Rueping, M. Nickel-catalyzed

c-heteroatom cross-coupling reactions under mild conditions via facilitated reductive elimination. _Angew. Chem. Int. Ed._ 60, 17810–17831 (2021). Article  CAS  Google Scholar  * Welin, E.

R., Le, C., Arias-Rotondo, D. M., McCusker, J. K. & MacMillan, D. W. C. Photosensitized, energy transfer-mediated organometallic catalysis through electronically excited nickel(II).

_Science_ 355, 380–385 (2017). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Tasker, S. Z., Standley, E. A. & Jamison, T. F. Recent advances in homogeneous nickel

catalysis. _Nature_ 509, 299–309 (2014). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Zhu, C., Yue, H., Nikolaienko, P. & Rueping, M. Merging electrolysis and

nickelcatalysis in redox neutralcross-coupling reactions: experiment and computation for electrochemically induced C–P and C–Se bonds formation. _CCS Chem._ 2, 179–190 (2020). Article  CAS 

Google Scholar  * Corcoran, E. B. et al. Aryl amination using ligand-free Ni(II) salts and photoredox catalysis. _Science_ 353, 279–283 (2016). Article  ADS  CAS  PubMed  PubMed Central 

Google Scholar  * Kudisch, M., Lim, C.-H., Thordarson, P. & Miyake, G. M. Energy transfer to Ni-amine complexes in dual catalytic, light-driven C–N cross-coupling reactions. _J. Am.

Chem. Soc._ 141, 19479–19486 (2019). Article  CAS  PubMed  PubMed Central  Google Scholar  * Das, S. et al. Photocatalytic (Het)arylation of C(sp3)–H bonds with carbon nitride. _ACS Catal._

11, 1593–1603 (2021). Article  CAS  Google Scholar  * Wang, K. et al. Covalent triazine frameworks via a low-temperature polycondensation approach. _Angew. Chem. Int Ed._ 56, 14149–14153

(2017). Article  CAS  Google Scholar  Download references ACKNOWLEDGEMENTS Y.Z. acknowledges China Scholarship Council (CSC) for providing a scholarship (CSC No. 201906280093). M.A. and A.S.

gratefully acknowledge Max Planck Society for the financial support of this project. J.S. acknowledges the National Natural Science Foundation of China (21972110). P.V.D.V., P.G.D., and

S.A. acknowledge the UGent concerted action grant BOFGOA2017000303, Ghent University BOF doctoral grant 01D04318 and the Research Foundation Flanders (FWO-Vlaanderen) grant no. G000117N.

Authors acknowledge Miss Jiamei Liu at Instrument Analysis Center of Xi’an Jiaotong University for her assistance with XPS analysis. Authors acknowledge Katharina ten Brummelhuis for the

low-temperature EPR spectra measurements; Dr. Nadezda V. Tarakina and Dr. Tobias Heil for TEM analysis. FUNDING Open Access funding enabled and organized by Projekt DEAL. AUTHOR INFORMATION

AUTHORS AND AFFILIATIONS * Department of Colloid Chemistry, Max Planck Institute of Colloids and Interfaces, 14476, Potsdam, Germany Yajun Zou, Stefano Mazzanti, Markus Antonietti & 

Aleksandr Savateev * State Key Laboratory of Electrical Insulation and Power Equipment, Center of Nanomaterials for Renewable Energy, School of Electrical Engineering, Xi’an Jiaotong

University, Xi’an, 710049, China Yajun Zou & Jian-Wen Shi * Center for Ordered Materials, Organometallics and Catalysis, Ghent University, 9000, Gent, Belgium Sara Abednatanzi, Parviz

Gohari Derakhshandeh & Pascal Van Der Voort * Department of Chemistry and Pharmacy Interdisciplinary Center for Molecular Materials (ICMM), Friedrich-Alexander University

Erlangen-Nürnberg, 91058, Erlangen, Germany Christoph M. Schüßlbauer & Dirk M. Guldi * Department of Inorganic Chemistry, Fritz-Haber-Institut der Max-Planck-Gesellschaft, Berlin, 14195,

Germany Daniel Cruz * Department of Heterogeneous Reactions, Max Planck Institute for Chemical Energy Conversion, Mülheiman der Ruhr, 45470, Germany Daniel Cruz Authors * Yajun Zou View

author publications You can also search for this author inPubMed Google Scholar * Sara Abednatanzi View author publications You can also search for this author inPubMed Google Scholar *

Parviz Gohari Derakhshandeh View author publications You can also search for this author inPubMed Google Scholar * Stefano Mazzanti View author publications You can also search for this

author inPubMed Google Scholar * Christoph M. Schüßlbauer View author publications You can also search for this author inPubMed Google Scholar * Daniel Cruz View author publications You can

also search for this author inPubMed Google Scholar * Pascal Van Der Voort View author publications You can also search for this author inPubMed Google Scholar * Jian-Wen Shi View author

publications You can also search for this author inPubMed Google Scholar * Markus Antonietti View author publications You can also search for this author inPubMed Google Scholar * Dirk M.

Guldi View author publications You can also search for this author inPubMed Google Scholar * Aleksandr Savateev View author publications You can also search for this author inPubMed Google

Scholar CONTRIBUTIONS Y.Z. characterization of CTFs, screening of CTFs in oxidative halogenation, and manuscript writing; S.A. synthesis of CTFs, manuscript writing; P.G.D. synthesis of

CTFs, manuscript writing; S.M. characterization of CTFs, dual Ni photocatalytic C‒N cross coupling; C.M.S. TAS study; D.C. UPS study; P.V.D.V. manuscript writing; J.W.S. manuscript writing;

M.A. manuscript writing; D.M.G. TAS study, manuscript writing; A.S. idea, coordination of the project, manuscript writing. CORRESPONDING AUTHOR Correspondence to Aleksandr Savateev. ETHICS

DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. PEER REVIEW PEER REVIEW INFORMATION _Nature Communications_ thanks Chenliang Su and the other, anonymous, and

reviewer for their contribution to the peer review of this work. Peer reviewer reports are available. ADDITIONAL INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to

jurisdictional claims in published maps and institutional affiliations. SUPPLEMENTARY INFORMATION SUPPLEMENTARY INFORMATION PEER REVIEW FILE RIGHTS AND PERMISSIONS OPEN ACCESS This article

is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give

appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in

this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative

Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a

copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Zou, Y., Abednatanzi, S., Gohari Derakhshandeh, P. _et

al._ Red edge effect and chromoselective photocatalysis with amorphous covalent triazine-based frameworks. _Nat Commun_ 13, 2171 (2022). https://doi.org/10.1038/s41467-022-29781-9 Download

citation * Received: 25 February 2021 * Accepted: 14 March 2022 * Published: 21 April 2022 * DOI: https://doi.org/10.1038/s41467-022-29781-9 SHARE THIS ARTICLE Anyone you share the following

link with will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature

SharedIt content-sharing initiative