Accelerated Direct Hydroxylation of Aryl Chlorides with Water to Phenols via Proximity Effect in Heterogeneous Metallaphotocatalyst

doi:10.21203/rs.3.rs-927401/v1

Download PDF

Article

Accelerated Direct Hydroxylation of Aryl Chlorides with Water to Phenols via Proximity Effect in Heterogeneous Metallaphotocatalyst

https://doi.org/10.21203/rs.3.rs-927401/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The proximity of different active sites in multi-catalytic systems is crucial in influencing the catalytic reactions, i.e., to occur or to be accelerated. Here we reported a heterogeneous metallaphotocatalyst prepared by embedding Ni(II) species in a photosensitive covalent organic framework (COF). A concerted triad catalytic behavior executed by electron, hole, and Ni species triggered a dramatic catalytic enhancement on the activation of aryl chlorides with water to phenols. It demonstrated a 50-fold increment in its activity when compared to homogeneous analog, e.g., NidabpyBr₂. Also, it was able to tolerate a broad range of functional groups even for those hardly achieved by the thermal catalysts. Based on the detailed study, the proximity of the photoactive COF and Ni(II) sites facilitates rapid electron transfer to produce Ni(I) active sites for the oxidative addition. Meanwhile, the photogenerated holes oxidized water to hydroxyl radical, which then attacked Ni(III) intermediates to complete the catalytic cycle.

Physical sciences/Chemistry/Catalysis/Photocatalysis

Physical sciences/Materials science/Materials for energy and catalysis/Photocatalysis

Physical sciences/Chemistry/Catalysis/Heterogeneous catalysis

Catalysis has exerted transformative influences on various fields that ranged from the production of chemicals, materials, and fuels to environment, health, and aerospace¹. Most of these well-established catalytic processes typically utilize catalysts to enable the proceeding of a single reaction, and this requirement has triggered intrinsic limitations on their efficiency and economy². Accordingly, to address the above-mentioned drawback, a multi-catalytic system is an ideal system that can synergistically promote a reaction since it enables transformations that are initially inaccessible when using a single catalyst. In addition, multi-catalytic system is poised to significantly reduce the time, waste, and cost of the reaction, which further increases its appeal³_. Recently, combining photoredox catalysis and alternative modes of catalysis has emerged as a powerful chemical synthesis tool^4,5. In this context, the integration of visible-light photocatalysis and transition metal catalysis, i.e., termed as metallaphotocatalysis, has been proven to be particularly advantageous in facilitating challenging carbon-carbon and carbon-heteroatom bond formations. This is because visible light can accelerate the otherwise sluggish or inaccessible elementary organometallic steps such as transmetalation, oxidative addition, and reductive elimination^6,7. In addition, functional group tolerance can be significantly improved and side reactions will be limited under mild reaction conditions, which are important challenges to address in thermal-driven catalysis⁸. Based on this cooperative catalytic system, photogenerated charge must diffuse from the photosensitizer to the reactants or metal active sites to initiate the redox process⁹.This, in turn, suggests that when the separation of the electrons and metal species or the separation of the holes and sacrificial agent is too large, concentration gradient will emerge and the catalytic activity will decrease¹⁰. It is worth noting that it is challenging to control the separation of these species in high entropy homogeneous systems¹¹. Moreover, the fast charge kinetics (normally from femtoseconds to microseconds) and slow reaction kinetics (normally from microseconds to seconds) can result in the low utilization of photogenerated charges, which can further limit the efficiency of the metallaphotocatalysis¹². Even though few studies have focused on achieving more efficient utilization of photogenerated charges by integrating the chemical tethering photosensitizer and redox catalyst into one molecule¹³ or co-incorporating photoactive metal complexes, e.g., iridium complex, and metal compounds, e.g., nickel salt, in porous materials^14-17, the development of heterogeneous metallaphotocatalysts with nanoscale intimacy between photosensitizer and transition metal catalyst remains scarce. Thus, to develop efficient metallaphotocatalysts, facilitating this nanoscale intimacy between photosensitizer and transition metal catalyst is urgently needed¹⁸.

Covalent organic frameworks (COFs) are crystalline porous materials that are constructed with covalently linked organic molecules. The implementation of reticular chemistry has rendered them an emerging class of photoactive semiconductors. COFs possess unique advantages such as tunable band gap and band edge, which make them highly attractive. Also, the periodic columnar π-π stacks can improve the charge transfer efficiency¹⁹. More importantly, the catalytic active species can be embedded within the COFs skeleton, which provides effective control of the distribution and coordination environment of the active sites and produces a significant synergistic effect between the support and the active sites^20,21. However, to date, reports on the acceleration of reaction via a synergistic effect from proximity orientation of metal active sites in COFs remain lacking, and this gap must be addressed.

Here we reported an elaborate system to control the intimacy between the photoactive COF and metal species at the nanoscale level, whereby significant positive effects on the catalytic efficiency during the activation of aryl chlorides were achieved. This system was realized with a heterogenous metallaphotocatalyst, i.e., Ni(II)-bpy-COF, that was prepared via a facile post-functionalization of Ni(II) sites on triazine-bipyridine-based COF. Upon visible light irradiation, the as-prepared catalyst could promote the direct hydroxylation of a wide range of aryl chlorides and water under mild reaction conditions to phenols with high yields. Interestingly, the catalyst was effective even for base-sensitive substituent or chlorobenzene bearing NH or NH₂ groups. It is worth noting that the as-prepared catalyst could demonstrate about 50-fold and 4-fold enhancement over the homogeneous analog and the physical mixture of NiBr₂·glyme and bpy-COF, respectively. Based on the experimental and theoretical analyses, the photogenerated electrons, holes, and Ni species were jointly responsible for this excellent performance exhibited by the catalyst. Also, the near-zero transfer distance between COF and Ni(II) helped to realize the rapid C-O bond formation rate.

Synthesis and characterization of catalyst. 2,2’-bipyridine-functionalized triazine-based COF (denoted as bpy-COF) was prepared under solvothermal conditions in a sealed glass tube by condensing 4’4’4’-(1,3,5-triazine-2,4,6-triyl)trianiline (TTA) and 5,5’-dialdehyde-2,2’-bipyridine (dabpy)²². Then, Ni ions were introduced into bpy-COF scaffold via a facile post-grafting treatment with NiBr₂·glyme, and the product was denoted as Ni(II)-bpy-COF. Ni(II) loading could be easily controlled by adjusting the amount of NiBr₂·glyme added (Fig. 1a). Based on the inductively coupled plasma optical emission spectrometer (ICP-OES) analysis, Ni contents in the series of Ni(II)-bpy-COF prepared in this work ranged between 0.83 wt.% to 4.62 wt.%, which indicates the presence of Ni species (Supplementary Table 1). Next, X-ray photoelectron spectroscopy (XPS) results revealed a peak located at 855.6 eV in the Ni 2p XPS spectrum, which indicates the presence of Ni²⁺ in all Ni(II)-bpy-COF samples (Supplementary Fig. 1)²³. To further characterize the coordination environment of the Ni atoms, X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) were performed by selecting a representative sample, i.e., Ni(II)-bpy-COF-5 with 3.44 wt.% loading. The Ni K-edge XANES spectra of NiO, NiBr₂·glyme, NidabpyBr₂, and Ni(II)-bpy-COF-5 are shown in Fig. 1b, whereby two peaks were observed for all samples. For Ni(II)-bpy-COF-5, a small pre-peak located at about 8332.2 eV could be assigned to a 1s-3d electron transition, while the second major peak located at about 8348.6 eV indicated the presence of Ni(II) species²⁴. As expected, the position of the pre-peak revealed a strong resemblance to the NidabpyBr₂ complex. However, it was different with NiO (8332.7 eV) and NiBr₂·glyme (8332.1 eV), which indicates the Ni(II) species in Ni(II)-bpy-COF-5 with NidabpyBr₂were in a similar chemical configuration_. Also, the Fourier-transformed k3-weighted EXAFS (Fig. 1c) image showed one main peak at about 1.61 Å for Ni(II)-bpy-COF-5, which correspond to the first coordination shell of Ni-N. When compared to the Ni foil, no Ni-Ni coordination peaks were observed. Such a result indicates that the atomically dispersed Ni was coordinated with N in COF. According to the EXAFS fitting results, the radius of Ni-N and Ni-Br were 2.05 and 2.55 Å (Supplementary Table 2), respectively, and the data could be well-fitted to the NidabpyBr₂model complex. These results demonstrate that a single Ni atom in COF had adopted a Ni-N₂Br₂ molecular configuration with an octahedral coordination environment¹⁵. As shown in Fig. 1d, Ni nanoparticles were not observed under the transmission electron microscope (TEM). The energy-dispersive X-ray (EDX) mapping images (inset) indicated homogeneous distributions of C, N, Br, and Ni elements within the COF matrix. The single Ni site in COF was confirmed by aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM)²⁵, and this is consistent with the EXAFS result (Fig. 1e). Furthermore, powder X-ray diffraction (PXRD) patterns (Supplementary Fig. 2) of bpy-COF and Ni(II)-bpy-COF-5 revealed the same set of diffraction peaks, which suggests the retention of high crystallinity after coordinating with Ni(II) species. The PXRD pattern of Ni(II)-bpy-COF-5 corresponded to the simulated 2D model based on an AA stacking structure in the hexagonal space group, among which the most intense peak located at 2θ of 2.54 correspond to the (100) plane refection²².

Solid-state UV-Vis diffuse reflectance spectra confirmed that both bpy-COF and Ni(II)-bpy-COF-5 absorbed light in the UV and visible light regions (up to 600 nm). It is worth noting that Ni(II)-bpy-COF-5 showed an intense and red-shifted light absorption between 450 to 550 nm (Fig. 2a). This result may be attributed to the extended electron delocalization in COF networks due to the embedded nickel ions, and the activation of π to π* electron transition arising from the strong interaction between nickel ions and COF networks²². Accordingly, the optical band gap of Ni(II)-bpy-COF-5 calculated from the Tauc plot was 2.54 eV, which was lower than that of bpy-COF (2.64 eV) (Fig. 2b). Furthermore, the positive slope of the plots obtained at different frequencies during the Mott-Schottky (MS) measurement indicated that these materials were n-type semiconductors. Since the bottom of the conduction band (LUMO) in n-type semiconductor is typically close to the flat-band potential, the LUMO locations of bpy-COF and Ni(II)-bpy-COF-5 were estimated to be -0.56 V vs. SCE and -0.42 V vs. SCE, respectively, by extrapolating the MS plot (Fig. 2c). Based on their corresponding band gaps, the valence band (HOMO) locations were determined to be at +2.08 V vs. SCE and +2.12 V vs. SCE (Fig. 2d) for bpy-COF and Ni(II)-bpy-COF-5, respectively. The negative shift in the conduction band edge of Ni(II)-bpy-COF-5 relative to that of bpy-COF was attributed to the delocalized intramolecular charge transfer (ICT) from COFs to the coordinated Ni(II) species. Moreover, considering the highly positive potential of HOMO in Ni(II)-bpy-COF-5, H₂O could be theoretically oxidized to hydroxyl radicals (OH·) under basic conditions upon photon excitation²⁶. Overall, Ni(II)-bpy-COF-5 demonstrated promising potential in promoting visible-light-induced photocatalytic reactions

Catalytic studies. Phenols, hydroxylated heteroarenes and their derivatives are widely used as key intermediates in organic synthesis, materials science, and drug discovery, with global production reaching more than 10 million tons per year²⁷. Despite the tremendous progress in the synthesis of phenols over the past decade, the current approaches usually suffer from low efficiency and high energy consumption, e.g., the overall yield of the Hock process is lower than 5%²⁸. To date, the transformation of aryl halides with water to phenols has emerged as an attractive synthetic strategy, whereby such reaction is typically accomplished by using aryl bromides or iodides²⁹. It is worth noting that aryl chlorides are more abundant, stable, and lower in cost than aryl bromides or iodides, which can serve as excellent candidates for this reaction. However, these chemicals are largely under-explored due to the need to overcome the high energetic barrier for the activation of C(sp²)-Cl bond (PhCl at ∼97 kcal/mol)³⁰.

To address this challenge, 4-chlorobenzonitrile and H₂O were first employed as the model substances to evaluate the potential of Ni(II)-bpy-COF during the visible-light driven hydroxylation. During the preliminary attempt, a reaction mixture comprising of 4-chlorobenzonitrile (0.50 mmol), H₂O (20 mmol), 2.0 mol% Ni(II)-bpy-COF-5, and triethylamine (0.75 mmol) in DMF/CH₃CN (5.0 mL, Vol. 1/1) was irradiated with 50W blue LEDs (440 nm) at 25^oC under N₂ atmosphere. After 12 hours, the corresponding C-O cross-coupled product 4-hydroxybenzonitrile with a moderate yield of 57.4% was achieved. The feasibility of this protocol was further investigated by conducting a series of control experiments. No reaction occurred in the absence of light, absence of triethylamine and H₂O, or absence of Ni(II)-bpy-COF (Table 1, entries 2-5). When NiBr₂·glyme was employed instead of Ni(II)-bpy-COF, no desirable product was achieved, and this result suggests that the photosensitive bipyridine functionalized-COF scaffold was essential to this reaction (Table 1, entry 6). When bpy-COF was employed instead of Ni(II)-bpy-COF, no desirable product was detected, which suggests the indispensable role of Ni in this cross-coupling reaction (Table 1, entry 7). Based on these results, it is clear that both photosensitive bipyridine-based COF scaffold and Ni(II) active species and the presence of visible light were necessary for this reaction to proceed³¹.

Table 1 Optimization of the reaction conditions.^a

Next, the effects of the various catalysts with different Ni loadings, catalyst loadings, different bases, H₂O and solvent amounts, reaction durations and temperatures on the efficiency of photocatalytic reaction were investigated. It was shown that the yield of 4-hydroxybenzonitrile increased with increasing catalyst loadings, whereby the yield reached a maximum at a catalyst loading of 3.44% (Supplementary Table 3). Further increase in the catalyst loading led to a decrease in the yield since excessive Ni(II) species inside the pores of COF could result in a greater mass transfer limitation. Meanwhile, as the catalyst loading increased from 2.0 mol% to 3.0 mol%, a reduced activity was recorded, and this could be attributed to the detrimental effects of excess solid powder used in the reaction towards light absorption (Supplementary Table 4). When compared to the thermally catalyzed hydroxylation reaction (must be performed under a strongly alkaline environment), this reaction could be triggered efficiently under a mildly alkaline environment. As shown in Supplementary Table 5, Et₃N exhibited the best performance among the other tests. In contrast, DIPEA, TEOA, and 2,6-lutidine only exhibited 31.0 %, 2.95%, and trace yields of the desired product, respectively. These results may be due to their lower dissociation constant values (pKa) between 5.30 to 10.98 (Supplementary Table 5, entry 1-4) as compared to Et₃N, which is unconducive towards the deprotonation of water. Moreover, inorganic bases such as K₂CO₃ could not effectively drive the hydroxylation reaction, which may be attributed to the inability of the inorganic base in capturing the holes. As such, photo-generated electrons and holes could be easily recombined in Ni(II)-bpy-COF-5. Meanwhile, the effects of Et₃N content on the product yield were also investigated, and the results showed that 4.0 equiv. was required to achieve a high yield of 82.8% (Supplementary Table 6). A series of organic solvents such as dimethyl sulfoxide (DMSO), N,N-dimethylformamide (DMF), toluene, dioxane, tetrahydrofuran (THF), acetonitrile (CH₃CN), and methanol (CH₃OH) were also studied. Based on the results, the mixture solvent containing 3.0 mL DMF and CH₃CN gave the highest yield since both were mutually dissolvable in H₂O and organic substrate (Tables S7-8). Also, the effects of H₂O reactant content on the product yield were investigated, and it was shown that H₂O was necessary to trigger this reaction and the optimum H₂O content was 360 µL (Supplementary Table 9). Furthermore, as the reaction duration increased from 12h to 24 h at 40^oC, the yield increased from 57.4% to 99% (Table 1, entry 1). Based on these results, an optimized reaction conditions with 2.0 mol% Ni(II)-bpy-COF-5, 40 equiv. H₂O, 4.0 equiv. Et₃N as the base, 3.0 mL DMF/CH₃CN as a solvent, and a reaction time of 24 h at 40^oC can be concluded. It is worth noting that the as-prepared catalyst was notably superior as compared to those previously reported systems that utilized metal-based complexes or photocatalysts. This is because the as-prepared catalyst was able to achieve the highest yield under mild reaction conditions without the use of strong bases and high-cost hydroxide resources (Supplementary Table 10).

Furthermore, the scope of substrates was extended to examine the applicability of Ni(II)-bpy-COF-5 catalyst in direct hydroxylation (Fig. 3). Chloroarenes with either electron-donating or electron-withdrawing groups, e.g., nitro, ethanoyl, trifluoromethyl, methoxy, hydroxyl, and methyl, were able to react smoothly to generate aromatic phenols 2a-2f with a yield range between 59%-99%. Moreover, heteroaryl chlorides such as 6-chloropyridine-3-carbonitrile were successfully coupled with H₂O, and the desired product 2g was achieved with a yield of 89%. Interestingly, various base-sensitive substituents on aryl chlorides were well-tolerated under mild conditions. For example, chlorobenzenes bearing ester or cyano groups could generate corresponding phenols 2h and 2i with a yield of 97% and 99%, respectively. Notably, the reaction worked well with substrates with NH or NH₂ groups, which enabled the convenient hydroxylation of such substrates in a protective group-free manner³². For instance, 6-chloro indole and 4-chloroaniline delivered the corresponding 6- and 3-hydroxyquinoline 2j and 2k with a yield of 59% and 79%, respectively. Thus, the advantage of our protocol lies not only in the ability to activate inert chloroarenes but also in significantly enhancing functional group tolerance when compared to the thermal catalyzed hydroxylation conditions. To illustrate the superiority of this work, the comparison was made with the previous literature and summarized in Supplementary Table 10.

Another important advantage of the practical potential of heterogeneous catalysts is recyclability³³. As shown in Supplementary Fig. 3, the catalyst could be recovered and reused for at least five cycles without significant loss in its catalytic performance between 4-chlorobenzonitrile and H₂O. According to the ICP analysis, the content of Ni species in the solution was less than 0.50 ppm after using the catalyst five times repeatedly. Such a result indicates that operating the catalyst under mild reaction conditions could effectively inhibit Ni(II) leaching from COF skeleton. XPS spectrum of the used Ni(II)-bpy-COF-5 revealed that the Ni 2p peak located at 855 eV remained unchanged (Supplementary Fig. 4). In addition, no apparent change in the XRD pattern was observed (Supplementary Fig. 5). Thus, based on the collective results, the composition and structure of Ni(II)-bpy-COF-5 were preserved after the cyclic test³⁴.

Determination of the role of COF. As indicated in the activity test, the reaction could not occur when NiBr₂·glyme was used as the catalyst without COF (Table 1, entry 6). To gain insights into the critical role of COF, trapping experiments were conducted by employing 4-chlorobenzonitrile and H₂O as the reactants. As shown in Fig. 4a, isopropanol, p-benzoquinone, AgNO₃, and Na₂S/Na₂SO₃ were used as the scavengers of hydroxyl radical (·OH), superoxide radical (·O₂^-), electron (e^-), and hole (h⁺), respectively. It was observed that the catalytic activity of the catalyst decreased dramatically after the addition of AgNaO₃. Also, adding isopropanol to the system caused the decline of product yield. In contrast, after adding p-benzoquinone or Na₂S/Na₂SO₃, a negligible decrease in the catalytic efficiency was observed. These results confirmed that photogenerated electrons played a dominant role in the reaction, and hydroxyl radical was also another contributing factor³⁴. Furthermore, in-situ electron paramagnetic resonance (EPR) spin trapping was employed to detect the generated hydroxyl radical (Fig. 4b). 0.50 mmol 4-chlorobenzonitrile, 20 mmol H₂O, 2.0 mol% Ni(II)-bpy-COF-5, 2.0 mmol Et₃N, 3.0 mL DMF/CH₃CN, and 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as radical trap were mixed and stirred for 5.0 minutes. Then, a small amount of the mixture was transferred into a capillary. Based on the result, no radical signal was observed under visible light illumination. Furthermore, a similar phenomenon was also observed for the system without light irradiation, base, catalyst, or water. However, a signal with g=3460 was obtained by removing 4-chlorobenzonitrile from the mixture under light irradiation. Such a result indicates the production of hydroxyl radicals via water oxidation in basic conditions by the photogenerated holes, and these holes would react with the reactant rapidly. Such a result was confirmed by the in-situ FTIR spectra, whereby rapid formation of hydroxyl products under the optimal reaction conditions was observed (Supplementary Fig. 6). Meanwhile, this observation eliminated the possibility of phenol radical generation during the reaction process. Thus, based on these results, both photogenerated electrons and holes from COF were simultaneously utilized, which could confer obvious benefits to the system such as the reduced likelihood of hole-electron recombination³⁵.

Analysis of Ni species redox cycle on COF. After studying the roles of COF, the next step was to elucidate the precise nature of the Ni-catalyzed C-O bond-forming process. Generally, two major plausible mechanisms are involved during the Ni-catalyzed cross-coupling reaction: (I) Ni(0)/Ni(II) cycle that proceeds with Ni(0)/Ni(II) oxidative addition and a subsequent Ni(II)/Ni(0) reductive elimination, and (II) Ni(I)/Ni(III) cycle that involves Ni(I)/Ni(III) oxidative addition and a subsequent Ni(III)/Ni(I) reductive elimination^36,37. It should be noted that using Ni(COD)₂@bpy-COF with Ni(0) species that was prepared by loading Ni(COD)₂ on bpy-COF, the resultant catalyst demonstrated significantly inferior performance with a product yield of 26.3% (Table 1, entry 9). Also, by employing a homogeneous NidabpyBr₂ complex, the as-prepared catalyst exhibited almost negligible catalytic efficiency with a product yield of 1.8% (Table 1, entry 8). Thus, we proposed that the reaction process was governed by Ni(I)/Ni(III) cycle, and Ni(I) active species was reduced by the photogenerated electrons³⁸. To determine the existence of electron transfer between the excited state of bpy-COF (marked as bpy-COF*) and Ni(II), steady-state emission quenching of bpy-COF* with varying NiBr₂·glyme concentrations was performed (Fig. 4c)³⁹. A weaker fluorescence intensity was observed at higher NiBr₂·glyme concentration. Meanwhile, the Stern-Volmer analysis exhibited an excellent linear regression, and the quenching efficiency was determined with a K_SV value of 2.329±0.2 M^-1 based on the Stern-Volmer equation (Supplementary Fig. 7). Furthermore, to explore the oxidation state of Ni during hydroxylation, electrochemical studies were performed on Ni(II)-bpy-COF-5 in MeCN. Fig. 4d shows the cyclic voltammetry (CV) curve of Ni(II)-bpy-COF-5 with two distinct irreversible reduction peaks located at -1.14 V and -1.87 V (vs. saturated calomel electrode (SCE)) in MeCN, which correspond to Ni^II/Ni^I and Ni⁰/Ni^0•‒ couples, respectively (green line). A similar reduction peak located at -1.09 V was also observed (orange line) when Et₃N was added, which can be ascribed to the base-stabilized Ni^I-species. However, only one reduction peak located at -1.28V was observed (blue line) in the CV curves of the individual bpy-COF sample, which corresponds to bpy-COF^0/•‒ couple. This result indicates the ability of bpy-COF to reduce Ni^II to Ni^I, but it could not further reduce Ni^I to Ni⁰.⁴⁰ Based on the collective results from steady-state emission quenching experiments and electrochemical tests, Ni^I-species generated during the reduction of electrons from the excited state of COF was the active species.

To better understand the impact of this Ni(I) catalyzed oxidative addition step on the efficiency of C-O coupling, Hammett plot was utilized to examine the rate dependencies on the reactants with substituents of varying electronic properties (Fig. 4e-4f, Supplementary Table 11). Specifically, the initial product formation rates of various electronically diverse 4-substituted aryl bromides were measured. Based on the plot, a correlation could be constructed, whereby a ρ value of +1.45 was calculated. This result confirms the significant role of the oxidative addition step in determining the overall C-O coupling rate. Next, density functional theory (DFT) calculations were performed to understand the reaction pathway of aryl halides with water as demonstrated in the experiments. As shown in Fig. 6a, during the first step, Et₃N cationic radical was generated via photogenerated hole oxidation, whereby the hydrogen atom transfer process occurred with water to produce hydroxyl radical. Such a step was viable since it possessed a driving force of 11.7 kcal mol^-1. Next, Ni^III species was produced from the oxidative addition of aromatic halides with Ni^I species (that was generated by photogenerated electrons), which possessed a barrier of 23.0 kcal mol^-1. Finally, OH radical directly underwent a concerted radical aromatic substitution reaction to form the desired product, i.e., phenol^40-42.

Furthermore, reaction kinetics were investigated to understand the overall reaction behavior of Ni(II)-bpy-COF-5 catalyst. Firstly, the initial rate was plotted against Ni loading (Supplementary Fig. 8). Interestingly, a roughly linear correlation was obtained at lower concentrations, which indicates the excessive photons under such conditions. As such, the nickel catalytic cycle was not limited by the accessibility to an excited photocatalytic species (photon-unlimited). Data obtained within this regime could facilitate the evaluation of the intrinsic kinetics of the nickel cycle¹¹. To probe the dependence of 4-bromobenzonitrile and water during hydroxylation (Supplementary Fig. 9-10), the relationship between the substrate concentration and initial rate was investigated. Based on the results, 4-bromobenzonitrile and water exhibited a positive order rate dependence (d(phenol)/d(t)=k(ArX)^1.4458_*(H₂O)^1.1869), which further demonstrated a concerted triad catalytic process facilitated by the electrons, holes, and Ni species in the Ni(II)-bpy-COF-5 catalyzed system⁴³.

Proximity effect of Ni(II)-bpy-COF. Three control catalysts with similar photo/Ni-dual catalytic systems, i.e., NidabpyBr₂/BODIPY, NiBr₂·glyme+bpy-COF, and NidabpyBr₂+g-C₃N₄, were examined (Fig. 5a). Under the same conditions, the yield of the reaction catalyzed by two homogeneous catalysts, i.e., NidabpyBr₂/BODIPY, was 23.4%. Also, the use of the physical mixture of homogenous NiBr₂·glyme and bpy-COF support resulted in a lower yield of 37.8%. Moreover, employing NidabpyBr₂+g-C₃N₄ as the catalyst led to a yield of 29.9% (Fig. 5b). These results may be attributed to the difference in the average distance between the catalyst and photosensitizer in each system. The reason for the reduced catalytic efficiency exhibited by NidabpyBr₂/BODIPY may be the inevitable interference of the charge transfer process by the surrounding organic solvents. NiBr₂·glyme+bpy-COF and NibpyBr₂+g-C₃N₄presented long transfer distance between the Ni catalyst that was soluble in the solvent and solid photosensitizer that was insoluble in the solvent⁴⁴.The lower values exhibited by NiBr₂·glyme+bpy-COF and NibpyBr₂+g-C₃N₄ when compared to that by Ni(II)-bpy-COF-5 may illustrate the need to achieve adequate proximity between the catalyst and photosensitizer for electron transfer. As a comparison, the distance between the photosenstive units and metal species was reduced significantly to near-zero since Ni(II) was directly coordinated with the bpy-groups in the COF framework. Moreover, due to the two-dimensional COF crystal structure, it could provide a high-speed charge transfer channel, which could also contribute to the excellent performances.

To compare the electron transfer efficiency between the excited states of different photo/Ni systems and aryl chlorides, Nidabpy+g-C₃N₄ and Ni(II)-bpy-COF-5 with varying 4-bromobenzonitrile concentrations were subjected to steady-state emission quenching experiments. As the concentration of 4-bromobenzonitrile increased, the fluorescence intensity grew weaker and the Stern-Volmer plot exhibited an excellent linear regression (Fig. 5c-e). The quenching efficiencies of Nidabpy/g-C₃N₄ and 3.44% Ni(II)-COF system were quantified by the Stern-Volmer equation, i.e., (I₀/I) = 1 + ksv[Q], which amounted in a quenching constant ksv of 0.35 and 6.10 for Nidabpy+g-C₃N₄ and Ni(II)-bpy-COF systems, respectively. As such, Ni(II)-bpy-COF was almost 17.4 times more effective than Nidabpy/g-C₃N₄ in luminescence quenched by 4-bromobenzonitrile. This result further confirmed the benefit of realizing the proximity between the Ni(II) species and COF to facilitate a more efficient charge transfer process, and this will then result in significantly improved catalytic performances⁴⁵.

Based on the collective results, the plausible reaction mechanism in the visible-light driven Ni(II)-bpy-COF catalyzed direct hydroxylation of aryl halides with water is proposed (Fig. 6b). Triazine-based bpy-COF is a photocatalyst, Ni(II) species act as the metal catalytic sites, and Et₃N captures the generated holes¹². Upon visible-light irradiation, photosensitive Ni(II)-bpy-COF is excited to produce an effective separation of the photogenerated electron-hole pairs. The photogenerated hole can trigger the oxidation of Et₃N to form a cationic radical. Then, the Et₃N radical undergoes a hydrogen atom transfer process with water to generate hydroxyl radical. During this stage, the photogenerated electron is utilized for the reduction of Ni^II species to Ni^I species via a single electron transfer. Then, the oxidative addition of aromatic halides to the Ni^I species should deliver the Ni^III species. Finally, the hydroxyl radical controls the concerted radical aromatic substitution reaction with Ni^III species, which leads to the formation of phenol to complete the catalytic cycle.

In summary, we have developed a facile method to incorporate Ni(II) species into the skeleton of a photoactive triazine-based bpy-COF. The resultant material was demonstrated as a novel heterogeneous metallaphotocatalyst, whereby it exhibited an unprecedented catalytic capability during the visible-light driven direct hydroxylation of aryl chlorides with water under mild conditions to form phenols. Based on the results, the spatial organization of single-site Ni catalyst in proximity with the COF support could facilitate the transfer of photogenerated electrons from COF to Ni centers. This could then accelerate the initial oxidation addition step in the catalytic cycle, whereas the holes would react with neighboring water to generate a highly active hydroxyl radical to execute the final aromatic substitution reaction. This work revealed for the first time that the proximity effect in heterogeneous metallaphotocatalyst is highly crucial, and it also highlighted the opportunity in designing multicatalytic photothermal catalysts for challenging organic transformations.

Synthesis of bpy-COF. A 10 mL glass pressure tube was charged with 4,4',4''-(1,3,5-triazine-2,4,6-triyl)trianiline (TTA) (92 mg) and 2,2-bipyridyl-5,5-dialdehyde (82 mg), mesitylene (5.1 mL), dioxane (0.90 mL), and glacial acetic acid (0.20 mL). The glass tube was flash-frozen in a liquid nitrogen bath, then it was subsequently degassed by performing three freeze-pump-thaw cycles and flame sealed under vacuum. The tube was allowed to warm back to room temperature, and it was later placed in an oven at 120°C and left undisturbed for 3.0 days. After which, the reaction mixture was cooled to room temperature, and a yellow solid was collected via centrifugation, and it was later washed repeatedly with THF to remove the trapped guest molecules. It was then Soxhlet extracted in THF for 24 h and dried under vacuum at 90°C for 12 h.

Synthesis of Ni(II)-bpy-COF. 50 mg bpy-COF and a certain amount of NiBr₂·glyme were dispersed in 20 mL acetonitrile. After refluxing the mixture for 8.0 h, the orange solid was filtered and washed thrice with acetonitrile to remove the uncoordinated Ni species. Then, it was dried under vacuum at 80^oC for 12 h to obtain the final product. Ni(II) loading was adjusted by controlling the amount of NiBr₂·glyme added in the initial mixture (between 0.025 to 0.20 mmol) to obtain the products denoted as Ni(II)-bpy-COF-1, Ni(II)-bpy-COF-2, Ni(II)-bpy-COF-3, Ni(II)-bpy-COF-4, Ni(II)-bpy-COF-5, and Ni(II)-bpy-COF-6, respectively.

Photocatalytic hydroxylation of aryl chlorides with water to phenols: In a typical run, oven-dried quartz vial (20 × 100 mm) equipped with a stir bar was charged with 4-bromobenzonitrile (91.1 mg, 0.5 mmol, 1.0 equiv.), and Ni(II)-bpy-COF-5 (2.0 mol% Ni). Then, DMF/CH₃CN (1:1, 3.0 mL), H₂O (360 µL, 40 equiv.), and Et₃N (2.0 mmol, 4.0 equiv.) were added and the vial was later sealed with a rubber plug. The reaction mixture was subsequently sonicated and stirred for 20 min. Thereafter, the mixture was degassed by bubbling nitrogen into the reaction system for 10 min. The reaction mixture was irradiated with Blue LED (50W) at 40°C under rapid stirring for 24 h. After which, the reaction mixture was filtered. The as-obtained organic phase was then diluted with chromatographic grade methanol and the volume was fixed in a 1.0 mL analysis bottle. The sample was analyzed by an Agilent 1260 Series Triple Quad high-performance liquid chromatography with Agilent C18 reverse-phase chromatographic column. The conversion of the aromatic halides and the yield of phenol products were determined by using trimethylbenzene as the internal standard sample. All data was repeated at least twice, and the data error was guaranteed within ± 5%.

Recyclability Test. To determine the recyclability of Ni(II)-bpy-COF-5, the sample was allowed to settle down after each run, and the clear supernatant liquid was decanted slowly. The residual solid catalyst was washed and reused with a fresh charge of solvent and reactants for the subsequent cycle performed under the same conditions.

Characterization. Powder X-ray diffraction (PXRD) spectra were obtained from a Rigaku Dmax-3C Cu-Ka. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) characterization and energy dispersive X-ray spectroscopy (EDS) was conducted on a FEI Theims Z system. An inductively coupled plasma optical emission spectrometer (ICP-OES, Varian VSTA-MPX) was utilized to determine the Ni content. Fourier transformation infrared spectra (FTIR) were obtained from an AVATAR 370 FT-IR spectrometer. The electronic states of Ni were analyzed by X-ray photoelectron spectroscopy (XPS) using a PerkinElmer PHI 5000CESCA system. All binding energies were calibrated based on the contaminant carbon (C1S = 284.6 eV) as the reference. UV-Vis diffuse reflectance spectra (DRS) were collected on a UV-Vis spectrophotometer (UV-Vis DRS, Shimadzu UV-2450), with BaSO₄ employed as a white standard. Spectrofluorometer (Varian Cary-Eclipse 500, Varian Medical Systems Inc, Palo Alto, CA) equipped with a 60 W xenon lamp (excitation source) was utilized to obtain the photoluminescence (PL) spectra. The widths of the slits for both excitation and emission were set at 5.0 nm.

X-ray absorption spectra. X-ray absorption fine structure (XAFS) spectra at Ni K (E₀ = 8333 eV) edge was performed at BL14W1 beamline of Shanghai Synchrotron Radiation Facility (SSRF) operated at 3.5 GeV under a “top-up” mode with a constant current of 240 mA. The XAFS data were recorded at 298 K under fluorescence mode with a Lytle ion chamber. The energy was calibrated accordingly to the absorption edge of pure Ni foil. Athena and Artemis codes were used to extract the data and fit the profiles. For the X-ray absorption near edge structure (XANES) result, the experimental absorption coefficients as a function of energy µ(E) were processed by subtracting the background and normalization. The result was reported as “normalized absorption” with E₀ = 8333 eV for all investigated samples and Ni foil/NiO standard. For the extended X-ray absorption fine structure (EXAFS) result, Fourier transformed (FT) data in R space were analyzed by applying a first-shell approximate model for Ni sample contribution. The parameters that described the electronic properties (such as the correction to the photoelectron energy origin, E₀), bond distance (R), and Debye-Waller factor around the absorbing atoms were varied during the fitting process.

Calculation method. Gaussian 09 software was used to perform all calculations. PBE0 functional⁴⁶ combined with D3BJ dispersion correction⁴⁷ was adopted in all calculations. For the optimization of geometry and frequency calculations, def2-SVP basis set⁴⁸ was employed. Singlet point energy calculations were performed with def2-TZVP basis set. SMD implicit solvation model⁴⁹ was used to account for the solvation effects in all calculations. The stability of wavefunction was examined for all compounds.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request. Source data are provided with this paper.

Acknowledgments

This work is supported by NSFC (22076122) and Shanghai government (19SG42, 19520710700 and 18230742500).

Author contributions

F. Z. conceived and designed the experiments. K. W. and H. J. performed the experiments along with the preparation of the supplementary materials. H. L. and H. C. helped to analyze the results of photocatalytic experiments. All authors discussed the results and helped in writing the manuscript.

Competing interests

The authors declare that they have no conflict of interest.

Lee, J. M. et al. Cooperative multi-catalyst systems for one-pot organic transformations. Chem. Soc. Rev. 33, 302–312 (2004).
Martínez, S. et al. Challenges and opportunities in multicatalysis. ACS Catal. 11, 3891–3915 (2021).
Kim, U. B. et al. Synergistic dual transition metal catalysis. Chem. Rev. 120, 13382–13433 (2020).
Nicewicz, D. A. et al. Merging photoredox catalysis with organocatalysis: the direct asymmetric alkylation of aldehydes. Science 322, 77–80 (2008).
Skubi, K. L. et al. Dual catalysis strategies in photochemical synthesis. Chem. Rev. 116, 10035–10074 (2016).
Prier, C. K. et al. Visible light photoredox catalysis with transition metal complexes: applications in organic synthesis. Chem. Rev. 113, 5322–5363 (2013).
Twilton, J. et al. The merger of transition metal and photocatalysis. Nat. Rev. Chem. 1, 0052 (2017).
Milligan, J. A. et al. Alkyl Carbon-Carbon Bond Formation by Nickel/Photoredox Cross-Coupling. Angew. Chem. Int. Ed. 58, 6152–6163 (2019).
Levin, M. D. et al. Photoredox catalysis unlocks single-electron elementary steps in transition metal catalyzed cross-coupling. ACS Cent. Sci. 2, 293–301 (2016).
Zecevic, J. et al. Nanoscale intimacy in bifunctional catalysts for selective conversion of hydrocarbons. Nature 528, 245–248 (2015).
Malik, J. A. et al. Evidence for photocatalyst involvement in oxidative additions of nickel-catalyzed carboxylate o-arylations. J. Am. Chem. Soc. 142, 11042–11049 (2020).
Till, N. A. et al. Mechanistic analysis of metallaphotoredox C-N coupling: photocatalysis initiates and perpetuates Ni(I)/Ni(III) coupling activity. J. Am. Chem. Soc. 142, 15830–15841 (2020).
Wang, D. et al. Development of tethered dual catalysts: synergy between photo- and transition metal catalysts for enhanced catalysis. Chem. Sci. 11, 6256–6267 (2020).
Feng, X. et al. Integration of earth-abundant photosensitizers and catalysts in metal-organic frameworks enhances photocatalytic aerobic oxidation. ACS Catal. 11, 1024–1032 (2021).
Zhu, Y. Y. et al. Merging photoredox and organometallic catalysts in a metal-organic framework significantly boosts photocatalytic activities. Angew. Chem. Int. Ed. 57, 14090–14094 (2018).
Quan, Y. et al. Bifunctional metal-organic layer with organic dyes and iron centers for synergistic photoredox catalysis. J. Am. Chem. Soc. 143, 3075–3080 (2021).
Reischauer, S. et al. Modular, self-assembling metallaphotocatalyst for cross-couplings using the full visible-light spectrum. ACS Catal. 10, 13269–13274 (2020).
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).
Wang, H. et al. Covalent organic framework photocatalysts: structures and applications. Chem. Soc. Rev. 49, 4135–4165 (2020).
Zhong, W. et al. A covalent organic framework bearing single Ni sites as a synergistic photocatalyst for selective photoreduction of CO₂ to CO. J. Am. Chem. Soc. 141, 7615–7621(2019).
Wang, X. et al. Covalent organic framework nanosheets embedding single cobalt sites for photocatalytic reduction of carbon dioxide. Chem. Mater. 32, 9107–9114 (2020).
Yang, S. et al. 2D covalent organic frameworks as intrinsic photocatalysts for visible light-driven CO₂ reduction. J. Am. Chem. Soc. 140, 14614–14618 (2018).
Wang, K. et al. Visible light-catalytic hydroxylation of aryl halides with water to phenols by carbon nitride and nickel complex cooperative catalysis. Green Chem. 22, 7417–7423 (2020).
Feth, M. P. et al. Investigation of the Ligand Exchange Behavior of Square-Planar Nickel(II) Complexes by X-ray Absorption Spectroscopy and X-ray Diffraction. Eur. J. Inorg. Chem. 838–847 (2003).
Chen, X. et al. Integrating single Ni sites into biomimetic networks of covalent organic frameworks for selective photoreduction of CO₂. Chem. Sci. 11, 6915–6922 (2020).
Luo, T. et al. Construction of C-C bonds via photoreductive coupling of ketones and aldehydes in the metal-organic-framework MFM-300(Cr). Nat. Commun. 12, 3583 (2021).
Xia, S. et al. Copper-catalyzed hydroxylation of (hetero)aryl halides under mild conditions. J. Am. Chem. Soc. 138, 13493–13496 (2016).
Fier, P. S. et al. Synthesis of complex phenols enabled by a rationally designed hydroxide surrogate. Angew. Chem. Int. Ed. 56, 4478–4482 (2017).
Yang, L. et al. Synthesis of Phenols: Organophotoredox/Nickel dual catalytic hydroxylation of aryl halides with water. Angew. Chem. Int. Ed. 57, 1968–1972 (2018).
Cui, B. et al. Defluorosilylation of fluoroarenes and fluoroalkanes. Nat. Commun. 9, 1–8 (2018).
Dong, W. et al. A highly stable all-in-one photocatalyst for aryl etherification: the NiII embedded covalent organic framework. Green Chem. 23, 5797–5805 (2021).
Song, Z. Q. et al. Palladium-catalyzed hydroxylation of aryl halides with boric acid. Org. Lett. 22, 8470–8474 (2020).
Qin, Y. et al. Solar-driven tandem photoredox nickel-catalysed cross-coupling using modified carbon nitride. Chem. Sci. 11, 7456–7461 (2020).
Filippini, G. et al. Light-driven, heterogeneous organocatalysts for C-C bond formation toward valuable perfluoroalkylated intermediates. Sci. Adv. 6, 9923 (2020).
Shang, W. et al. Synergistic redox reaction for value-added organic transformation via dual-functional photocatalytic systems. ACS Catal. 11, 4613–4632 (2021).
Gisbertz, S. et al. Overcoming limitations in dual photoredox/nickel-catalysed C-N cross-couplings due to catalyst deactivation. Nat. Catal. 3, 611–620 (2020).
Diccianni, J. B. et al. Mechanisms of nickel-catalyzed cross-coupling reactions. Trends in Chemistry 1, 830–844 (2019).
Zhu, C. et al. Nickel-catalyzed c-heteroatom cross-coupling reactions under mild conditions via facilitated reductive elimination. Angew. Chem. Int. Ed. 60, 17810–17831 (2021).
Tian, L. et al. Transient absorption spectroscopy offers mechanistic insights for an Iridium/Nickel-catalyzed C-O coupling. J. Am. Chem. Soc. 142, 4555–4559 (2020).
Zhu, B. et al. Ir(III)/Ni(II)-Metallaphotoredox catalysis: the oxidation state modulation mechanism versus the radical mechanism. Chem. Commun. 54, 5968–5971 (2018).
Ren, H. et al. How does Iridium(III) photocatalyst regulate Nickel(II) catalyst in metallaphotoredox-catalyzed C-S cross-coupling? theoretical and experimental insights. ACS Catal. 9, 3858–3865 (2019).
Yuan, M. et al. On the Nature of C(sp(3))-C(sp(2)) bond formation in Nickel-catalyzed tertiary radical cross-couplings: a case study of Ni/photoredox catalytic cross-coupling of alkyl radicals and aryl halides. J. Am. Chem. Soc. 142, 7225–7234 (2020).
Xia, Z. et al. Modifying electron transfer between photoredox and organocatalytic units via framework interpenetration for beta-carbonyl functionalization. Nat. Commun. 8, 361 (2017).
Ma, X. et al. Modulating coordination environment of single-atom catalysts and their proximity to photosensitive units for boosting MOF photocatalysis. J. Am. Chem. Soc. 143, 12220–12229 (2021).
De Abreu, M. et al. Synergistic photoredox/transition-metal catalysis for carbon-carbon bond formation reactions. Eur. J. Org. Chem. 1327–1378 (2020).
Adamo, C. et al. Toward reliable density functional methods without adjustable parameters: The PBE0 model. J. Chem. Phys. 110, 13 (1999).
Grimme, S. et al. A consistent and acc urate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132, 154104 (2010).
Weigend, F. et al. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. Phys. Chem. Chem. Phys. 7, 3297–3305 (2005).
Marenich, A. V. et al. Universal solvation model based on solute electron density and on a continuum model of the solvent defined by the bulk dielectric constant and atomic surface tensions. J. Phys. Chem. B 113, 6378–6396 (2009).

There is NO Competing Interest.

Supportinginformation.docx
Accelerated Direct Hydroxylation of Aryl Chlorides with Water to Phenols via Proximity Effect in Heterogeneous Metallaphotocatalyst

Download PDF

Version 1

posted

You are reading this latest preprint version

Accelerated Direct Hydroxylation of Aryl Chlorides with Water to Phenols via Proximity Effect in Heterogeneous Metallaphotocatalyst

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Conclusions

Methods

Data availability

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1