Consideration factors of electrode materials and pulsed electrochemical method
First, to achieve one-pot conversion of NO2− and arylboronic acids to arylamines, tandem NO2− electroreduction to NH3 followed by C − N coupling of NH3 with boronic acids should be well interconnected (Fig. 2a). The key is to develop a highly active electrode and catalyst for each step. Cu-based materials are prevailing electrocatalysts for NO2− reduction to NH3 due to their low cost and good catalytic activity.26 Tuning the electronic structure and/or coordinated surroundings can further enhance their electrochemical performances. We suppose that the low-coordinated Cu electrode, which can be readily produced via electroreduction of copper oxide or hydroxide precursor27–28 with abundant active sites to facilitate the adsorption of NO2− will be conducive to improving the selectivity and Faradic efficiency (FE) of NH3 (Fig. 2c). The efficient production of NH3 is highly significant to the next C − N coupling step with boronic acids. In addition, Cu catalysts are widely applied in the Chan-Lam coupling reactions.20–21 Therefore, Cu materials may be the ideal choice to implement the cascade transformation from NO2− to valuable organic amines. Second, Cu(II) is proven to be the active center for the Chan-Lam coupling reactions.20–21 Metallic Cu is easily oxidized under proper anodic potentials to leach Cu(II) into the electrolyte.29–31 We propose to produce Cu(II) via in situ electrooxidation of the Cu electrode rather than extra addition, which is favorable to lower the cost and simplify the operations. Furthermore, phenol, one of the main byproducts for the coupling of ammonia/amines with boronic acids in aqueous solutions,20–21 would be promoted by increasing the concentration of OH− ions.32–33 Although the pH value increases during NO2− electroreduction, we think that reverse electrooxidation will consume OH− ions due to electrode self-oxidation or the oxygen evolution reaction (OER),29–30,34−35 hence suppressing undesired phenol byproduct (Fig. 2d). However, continuous electrooxidation is unfavorable owing to the request of a catalytic amount of Cu(II) for the Chan-Lam coupling reactions and the easy oxidation nature of NH3/arylamines.36–37 Third, pulsed electrochemical methods (e.g., square potential waveforms) can induce alternate oxidation and reduction by controlling pulsed frequencies (Fig. 2b), which will alter the concentration of substrates or intermediates in the electrical double layer (EDL) to affect the product distributions.38–45 Based on these considerations, we propose that the pulsed technique will provide an ideal platform to realize the cascade conversion of NO2− to organic amines through sequential NO2− electroreduction over a Cu cathode and C − N coupling catalyzed by in situ dissolved Cu(II) by anode self-oxidation (Figs. 2b-e).
Synthesis And Characterization Of Low-coordinated Cu Nanocorals
We first synthesize self-supported Cu2O nanocorals (NCs) precursors on Cu foam via the thermal treatment of Cu(OH)2 nanowires (NWAs) under an Ar atmosphere (Supplementary Fig. 1 and Note 1). The scanning electron microscopy (SEM) images, X-ray diffraction (XRD) pattern, and X-ray photoelectron spectroscopy (XPS) spectra suggest the successful preparation of Cu2O NCs (Figs. 3a, c, and d). Then, LC-Cu NCs are fabricated through facile electroreduction of Cu2O NCs in 0.25 M phosphate-buffered saline (PBS, pH = 5.8) at − 0.8 V vs. Hg/HgO (Supplementary Fig. 2 and Note 2) (potentials in this work are all referred to as Hg/HgO unless stated otherwise). SEM images reveal the maintenance of the nanocoral morphology after electroreduction (Fig. 3a). However, the surface becomes rough, exposing more active sites, which can promote the adsorption of substrates and thus enhance the reaction efficiency. In situ Raman spectra show that the characteristic peaks of Cu2O at approximately 148, 220, and 625 cm− 1 vanish within 15 mins (Fig. 3b), suggesting the fast conversion of Cu2O to Cu. In situ XRD is also carried out to monitor the phase transformation of Cu2O. As depicted in Fig. 3c, after electrolysis for 15 mins, all the diffraction peaks in the XRD pattern are indexed to the pure Cu phase (JCPDS NO. 04-0836), corresponding to the in situ Raman results. In the XPS spectra (Fig. 3d), the peaks located at 952.2 and 932.3 eV belong to Cu0 2p1/2 and Cu0 2p3/2,26–27 respectively, which shift slightly toward low energy regions compared with those of Cu2O.
Furthermore, X-ray absorption spectroscopy (XAS) is used to determine the electronic structure and coordination environment of LC-Cu NCs. As shown in Fig. 3e, the Cu K-edge X-ray absorption near-edge structure (XANES) of LC-Cu exhibits similar features to that of the Cu foam, revealing that the reduced sample mainly consists of metallic Cu. In addition, the Fourier transformed k3-weighted Cu-K edge EXAFS spectra show the new appearance of the Cu − Cu path in the reduced sample (approximately 2.23 Å), while the average Cu − Cu coordination shell of LC-Cu is significantly lower than that of Cu foam (Fig. 3f). These results demonstrate that LC-Cu NCs formed via electroreduction of Cu2O NCs possess more low-coordination sites, which can be expected to facilitate electroreduction of NO2− with high performance.
One-pot C − n Coupling By Integrating Heterogeneous And Homogeneous Catalysis
NO2− electroreduction can proceed over the entire pH range,2 the HER is much easier in an acidic electrolyte than in neutral and alkaline media,46–47 and phenol byproduct is easy to form at a high concentration of OH− ions in the Chan-Lam coupling reactions.32–33 Moreover, methanol (MeOH) plays a critical role in inhibiting the HER and boosting the selectivity of NO3−/NO2− electroreduction.48 MeOH is also an ideal solvent in the coupling of boronic acids with NH3, as it can ensure a high concentration of NH3 in solution by the hydrogen bonding effect. Consequently, 0.25 M PBS electrolyte is selected as the electrolyte that mixes with MeOH to explore the proposed one-pot reactions. First, we examine the performance of NO2− electroreduction over the LC-Cu cathode. The linear sweep voltammetry (LSV) polarization curve displays a sharp increase in current density after adding 2.0 mmol of NaNO2 into the anodic solution (Fig. 4a), indicating a much easier reduction of NO2− than water. In comparison to Cu foam, LC-Cu NCs require a smaller overpotential to achieve a benchmark current density of 30 mA cm− 2 (896 vs. 756 mV, Fig. 4b), implying the high activity of Cu NCs toward NO2− electroreduction.
Next, we examine the feasibility of a one-pot transformation of NO2− and arylboronic acids to arylamines by using the pulsed electrochemical method that periodically applies an anodic potential (Ean) and a cathodic potential (Eca). Figure 4c reveals that − 1.1 V is optimal for NO2− electroreduction, giving rise to NH3 with the highest yield. We thus select Ean = − 1.1 V to screen other parameters of this proposed pulsed electrosythesis. Initially, 0.1 mmol of phenylboronic acid (1a, as the model substrate) and 2.0 mmol of NaNO2 are added into the cathodic cell that contains 0.25 M PBS electrolyte and MeOH. As a test, a 2 s pulse at Ean = 0.4 V is followed by a 2 s pulse at Eca = − 1.1 V (Supplementary Fig. 3 and Note 3), and the sequence is repeated for 5 h. After the reaction is finished, we detect the production of 0.2 mmol of NH3 from the reaction system (Supplementary Fig. 4). However, no phenylamine (2a) product is formed. Then, potentiostatic electrolysis was performed for 1 h before beginning the pulsed electrolysis as described above, and 2a was produced in an 11% isolated yield. It has been reported that rich NH3 is significant for achieving high yields and selectivity of arylamines in the reaction of arylboronic acids with NH3.32–33 That is, the accumulation of NH3 is necessary for the subsequent coupling of 1a with NH3. Thus, we conducted continuous electrolysis of 2.0 mmol of NO2− at − 1.1 V for 7 h before pulsed electrolysis. Nearly full conversion of NO2− is finished, and 1.80 mmol of NH3 is finally obtained. The massive formation of NH3 provides a guarantee for subsequent C − N coupling to arylamines.
Further screening results reveal that the conversion of 2a shows a volcanic trend with altering Ean and tan. A maximum 72% yield of 2a is obtained at Ean = 0.4 V, while tan = 2 s is the best interval for the anodic potentials (Figs. 4d and e). We attribute the inferior conversion of 2a at lower potentials (Ean < 0.4 V) and with a shorter tan (1.5 s) to the insufficient formation of Cu(II) under such conditions. Meanwhile, the concentration of Cu(II) ions reflects positive correlations to the applied anodic potentials (Fig. 4f), which may provide support for a higher 2a yield at a more positive potential. In addition, a further increase in Ean and tan is not helpful for the formation of 2a, which may be due to the increased consumption of NH3 and the electrooxidation of 2a at Ean = 0.6 V and with tan > 2.5 s. Indeed, the LSV curves of LC-Cu NCs demonstrate that NH3 and 2a are more easily oxidized than the OER, and a much lower potential is required for NH3 oxidation than that of 2a (Supplementary Fig. 5 and Note 4). Moreover, we observed a rapid color change of the solution and detected the azobenzene product by gas chromatography‒mass spectrometry (GC‒MS) when subjecting 2a to pulsed electrolysis with an Ean of 0.6 V (Supplementary Fig. 6 and Note 5), indicating the oxidation of 2a. Therefore, we identify the optimal pulsed conditions: Eca = − 1.1 V, Ean = 0.4 V, and tan = tca = 2 s.
Unveiling The Roles Of The Pulsed Potential Technique And Proposing A Possible Mechanism
Control experiments are carried out to further investigate the roles of the pulsed potential technique. First, Cu(II) is well known as the key catalytic species for the Cham-Lam coupling reactions. The ultraviolet‒visible (UV‒vis) fluorescence spectra show the presence of Cu(II) ions during the reaction (Fig. 5a), which is due to the electrooxidation-induced leaching of Cu(II) at the pulsed Ean. Then, the coordination of Cu(II) with NH3 is essential for triggering subsequent transformations.21 The formed Cu(II)-NH3 complex is further confirmed by electron paramagnetic resonance (EPR) tests (Fig. 5b), and a blue color is also displayed owing to the dissolution of the complex in the solution (Supplementary Fig. 7 and Note 6). Furthermore, in situ Raman spectra reveal that Cu2O can be formed at lower oxidation potentials and that CuO appears only when the potential reaches 0.2 V (Fig. 5c). Thus, higher oxidation potentials are required for the formation and leaching of hypervalent Cu(II) species to catalyze C − N couplings. This may account for why the yield of 2a increases as Ean increases until Ean = 0.4 V. Most importantly, Cu(II) can be readily removed from the solution by electroplating when the reaction is complete to ensure that the Cu level meets the emission standards before safe disposal (Supplementary Fig. 7 and Note 6). The expedient recycling of Cu(II) not only avoids the waste of Cu metals but also reduces the detriment of Cu residuals to the products and environment. Second, the pH value of the solution becomes 11.8 when the reaction is finished, which is lower than that of the one-pot NO2− to 2a by adding extra Cu(II) and under potentiostatic conditions (pH = 13.5, Fig. 5d and Supplementary Table 1). The much slower increase in pH in our system is mainly attributed to the electrooxidation of Cu (2Cu + 2OH− − 2e− → Cu2O + H2O, Cu2O + 2OH− − 2e− → 2CuO + H2O, Cu + 2OH− − e− → CuO + H2O)29–30 and ammonia at Ean,36 causing the consumption of OH−. As a result, the phenol byproduct is suppressed (Fig. 5d, Supplementary Fig. 8, and Supplementary Note 7). Third, compared to potentiostatic electrolysis at − 1.1 V, Cu(II) can remain stable for a longer period under pulsed conditions (Supplementary Fig. 9 and Supplementary Note 8), offering a great opportunity for achieving long-term electrolysis mediated by molecular metal catalysts. Fourth, arylboronic acid is easily activated by OH− ions to form nucleophilic PhB(OH)3−,49 which will migrate toward the anode during anodic time owing to electrostatic attraction. This will mitigate mass transport limitations and facilitate the interaction of PhB(OH)3− with the Cu(II)-NH3 complex, thereby speeding up C–N bond formation. Importantly, due to the rapid polarity switch, the concentrations of Cu(II) and NH3 are high near the electrode surface, thus promoting the Chan-Lam coupling. Finally, the overall reaction scheme is described in Fig. 5e. Electroreduction of NO2− first occurs over the surface of the LC-Cu cathode to generate NH3. When the potential is switched to Ean, Cu(II) is generated and then quickly coordinates with NH3 in the EDL to form the Cu(II)-NH3 intermediate. This intermediate further reacts with the activated PhB(OH)3− that migrates to the EDL to yield 2a and release the Cu(II) catalyst. Cu(II) either participates in the next catalytic cycle or is deposited on the electrode surface when applying a cathodic potential. Furthermore, the XRD patterns, XPS spectra, and Cu LMM Auger electron spectroscopy (AES) spectra of LC-Cu NCs reveal that copper oxides are formed on the surface of Cu after pulsed electrolysis (Supplementary Fig. 10 and Supplementary Note 9). The formed copper oxides can also be reduced to form Cu(0) to repeat the whole reaction process.
The methodology universality.
On the basis of the above understandings, we evaluate the generality of our method in the synthesis of functionalized primary arylamines from NO2− with different arylboronic acids. The − Me- and − OMe-substituted arylboronic acids and aryl Bpin are all amenable to our strategy, producing the corresponding amines in good yields. Delightedly, the yields of these products are all higher than those using the one-pot procedure by requiring an additional Cu(II) source and under constant potential conditions (Fig. 6a), showing the high efficiency of our method. Additionally, 15N-labeled arylamines, which have demonstrated potential applications in the preparation of 15N-labeled drugs for studying their metabolic profiles,50 can be facilely synthesized by using low-cost and easy-to-handle Na15NO2 as the 15N source (Fig. 6b), demonstrating more economic and safety advantages than using 15NH3. Furthermore, the pulsed electrochemical protocol can be expanded to the Click reactions of terminal alkynes with benzyl azides to construct 1,2,3-triazole N-heterocycles (Fig. 6c), implying good methodology universality.