Electrochemical Aziridination of Tetrasubstituted Alkenes with Ammonia

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Feb 2022Electrochemical Aziridination of Tetrasubstituted Alkenes with Ammonia Shuai Liu†, Wenxuan Zhao†, Jin Li†, Na Wu, Chang Liu, Xin Wang, Shuohao Li, Yan Zhu, Yong Liang and Xu Cheng Shuai Liu† Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 †S. Liu, W. Zhao, and J. Li contributed equally to this work.Google Scholar More articles by this author , Wenxuan Zhao† Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 †S. Liu, W. Zhao, and J. Li contributed equally to this work.Google Scholar More articles by this author , Jin Li† Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Jiangsu Provincial Engineering Laboratory of Advanced Materials for Salt Chemical Industry, College of Chemical Engineering, Huaiyin Institute of Technology, Huaian, Jiangsu 223003 †S. Liu, W. Zhao, and J. Li contributed equally to this work.Google Scholar More articles by this author , Na Wu Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Google Scholar More articles by this author , Chang Liu Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Google Scholar More articles by this author , Xin Wang Henan-Macquarie University Joint Center for Biomedical Innovation, School of Life Sciences, Henan University, Kaifeng, Henan 475004 Google Scholar More articles by this author , Shuohao Li Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Google Scholar More articles by this author , Yan Zhu Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Google Scholar More articles by this author , Yong Liang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Google Scholar More articles by this author and Xu Cheng *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 State Key Laboratory of Elemento-Organic Chemistry, Nankai University, Tianjin 300071 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202100826 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Ammonia (NH3) is an ideal nitrogen source in terms of availability, reactivity, safety, atom economy, environmental compatibility, and ease of isolation. However, its utility for amine synthesis is limited by its high bond dissociation energy, its strong coordination ability, and the difference between its reactivity and that of the product amines. Herein, we reported the first electrochemical protocol for direct syntheses of unprotected tetrasubstituted aziridines with NH3 and alkenes in the absence of an oxidant, which are highly challenging to achieve by other methods. The combination of graphite felt as the anode material and MeOH as the solvent was the key to the success of the protocol, and the effects of these factors were investigated by means of cyclic voltammetry and density functional theory calculations. Download figure Download PowerPoint Introduction Ammonia (NH3) is among the most widely used inorganic chemicals; global production in 2019 was estimated at 190 million metric tons (Figure 1a).1 NH3, which has a narrow flammable range (10%) in reaction,2 has been utilized as the ultimate nitrogen source in amine production because it contains 82% nitrogen and can readily be removed from reaction mixtures. Direct introduction of NH3 into molecules not only provides unprotected amines with the highest possible atom and step economy but also facilitates derivatization.3 Despite these advantages, direct functionalization of unpolarized molecules with NH3 poses several challenges (Figure 1b). The first is the strength of the N–H bond, which has a bond dissociation energy (BDE) of 99.5 kcal mol−14 and the second is the strong coordinating ability of the nitrogen atom, which results in the formation of stable amido hydride complexes, as revealed in the pioneering work by Hartwig’s group.5 Recent studies have addressed these two problems. For example, Chang’s group6 used a Cu(I) catalyst to avoid strong coordination of NH3 in chelation-assisted amination reactions of aryl C(sp2)–H bonds, and Wu and co-workers7 used a combination of photocatalysis and cobalt catalysis to synthesize aniline from NH3 and benzene. A third challenge originates from the fact that the primary amine products of amination reactions are more reactive than NH3 and thus undergo further reaction in situ. To overcome this challenge, the groups of Nicewicz8 and Lei9 utilized ammonium salts as NH3 surrogates in their novel aryl C(sp2)–H amination reactions. Despite these breakthroughs in C(sp2)–NH2 compounds with NH3, mild methods for the construction of C(sp3)–N products, which have even higher electron density, from NH3 and unpolarized substrates remain to be developed. Figure 1 | (a–f) Reaction design of aziridination using NH3 as the simplest nitrene equivalent. Download figure Download PowerPoint Aziridines, which are highly strained three-membered N-heterocycles, react with a variety of reagents to afford valuable amine derivatives.10 In addition, aziridines can serve as pharmaceutical intermediates,11 are present in various natural products,12 and can be used as building blocks for polymers (Figure 1c).13,14 Direct aziridination of alkenes is an efficient method for converting planar substrates to saturated N-heterocycles with three-dimensional structures (Figure 1d). The most extensively adopted methods for aziridination of alkenes require nitrogen sources with a protected nitrogen atom under transition-metal (TM) catalysis.15 Typically, the nitrogen source reacts with an oxidant to give a nitrene or nitrogen radical stabilized by a TM catalyst and then adds to the alkene, giving an aziridine that retains the protecting group. In seminal work on the aziridination of alkenes, Jat et al.16 addressed this issue using 2,4-dinitrophenyl hydroxylamine as a nitrogen source, which affords N–H aziridines in the absence of an external oxidant. In light of the new 12 green chemistry principles promulgated recently by Zimmerman et al.,17 it is still highly desirable to develop oxidant-free aziridination reactions that use the nitrene equivalent as simple as possible. Therefore, we set about to develop a method for aziridination of unpolarized alkenes with NH3 under catalyst- and oxidant-free conditions. The lone pair electrons on the nitrogen atom in aziridine can be inhibited from participating in conjugation by the rigid conformation, which may avoid the undesired further side reactions illustrated in Figure 1b. In the aziridination process, a driving force is necessary to facilitate the reaction of NH3 and alkenes under mild conditions. We speculated that an electrochemical process might be suitable for this purpose. Electrochemistry has emerged as a powerful tool for realizing difficult transformations,18–41 for example, intermolecular reactions to form congested ethers42 and constrained epoxides,43 as well as diazidation of alkenes44–46 and valuable functional materials.47–49 In addition, the formations of C–N bonds via dehydrogenative coupling were especially fulfilled by electrochemical approaches.50–70 In 2002, Yudin and co-workers71–73 reported the pioneering work on the electrochemical aziridination of alkenes with N-aminophthalimide as nitrogen source via nitrene pathways (Figure 1e). The following achievement was demonstrated by Little and co-workers74 via nitrogen radical pathway. In 2018, our group75 reported the aziridination with sulfamate involving two-electron oxidation of alkene. Very recently, Noël’s group76 reported a breakthrough in the aziridination using primary amine/NH3 of internal alkenes with a flow reaction. However, unprotected tetrasubstituted azirdine is still very hard to achieve. Herein, we report the aziridination reactions of tetrasubstituted alkenes with NH3 by means of an electrochemical approach in the absence of an external oxidant (Figure 1f). Experimental Methods A 10 mL three-necked heart-shaped flask was charged with the substrate alkene, Mg(ClO4)2 (0.1 mmol), and a magnetic stir bar. The flask was equipped with a rubber stopper, graphite felt (GF; 2 cm × 1 cm × 0.5 cm) as the anode, and a Pt plate (1 cm × 1 cm) as the cathode. The flask was evacuated and backfilled with NH3 gas three times, and then an NH3 gas balloon was connected to this flask via needle. Next, 5 mL of anhydrous MeOH was added via syringe. Electrolysis with constant cell potential was carried out in an ice-water bath. After completion of the reaction monitored with thin-layer chromatography (TLC) and gas chromatography mass spectrometry (GC-MS), the mixture was concentrated under reduced pressure. The residue was purified by chromatography on silica gel or neutral alumina to afford the desired product. More experimental details and characterization are available in the Supporting Information. Results and Discussion Reaction optimization The study began by carrying out reactions of tetrasubstituted alkene 1a as a model substrate because we expected that such an alkene would exert the greatest steric repulsion toward the approaching nitrogen source (Table 1; for more details, see Supporting Information Tables S1 and S3). Our first experiment was conducted in dimethylformamide (DMF) at room temperature under NH3 (1 atm) using electrolysis in an undivided cell with GF electrodes and LiClO4 as a supporting electrolyte. Disappointingly, under these conditions, 1a underwent hydrogenation to afford alkane 3a as the only appreciable product (62% isolated yield, entry 1). However, when MeOH was the solvent, the chemoselectivity was completely reversed, and desired product 2a was obtained in 46% NMR yield with no trace of 3a (entry 2). The yield of 2a increased to 70% when Mg(ClO4)2 was used as the supporting electrolyte and Ag as the cathode material with cooling to 0 °C (entry 3). The application of Pt cathode could further improve the isolated yield to 72% (entry 4). Interestingly, when Pt was used as the anode material, the reaction of 1a was completely shut down (entries 5 and 6). If EtOH was used as solvent, room temperature was required to achieve the dissolution of Mg(ClO4)2 and the NMR yield dropped to 46%. Finally, a reaction showed a more economical condition utilizing GF as both electrodes and LiCl as the supporting electrolyte (entry 8). An experiment using constant current electrolysis of 12 mA/cm2 gave the desired product 2a in 50% yield (entry 9). Table 1 | Reaction Conditions Optimizationa Entrya Solvent Electrode Supporting Electrolyte Yield of 2a Yield of 3a 1 DMF GF+/GF− LiClO4 Trace 65e(62f) 2b CH3OH GF+/GF− LiClO4 46 0 3c CH3OH GF+/Ag− Mg(ClO4)2 70(66) 0 4c CH3OH GF+/Pt− Mg(ClO4)2 74(72) 0 5b CH3OH Pt+/Pt− Mg(ClO4)2 N.R. N.R 6b CH3OH Pt+/GF− Mg(ClO4)2 N.R. N.R 7b EtOH GF+/Pt− Mg(ClO4)2 46 0 8d CH3OH GF+/GF− LiCl 70 0 9g CH3OH GF+/GF− Mg(ClO4)2 50 0 aConditions: 1a (0.2 mmol), supporting electrolyte (0.1 mmol), solvent (5.0 mL), 10 V cell potential, r.t., 4 h. b6.5 V cell potential, r.t., 4h. c6.5 V cell potential, 0 °C, 4 h. dLiCl (0.2 mmol), 6.5 V cell potential, 0 °C, 3 h. e1H NMR yield. fIsolated yield. gConstant current, 12 mA/cm2. Substrate scope Next, we explored the scope of the aziridination protocol by carrying out reactions of alkenes with various substitution patterns and functional groups (Figure 2). Heterocyclic aromatic substituents such as thiophene ( 2b, 61%) and pyridine ( 2c, 54%) were well tolerated. In addition, the fluorine and chlorine atoms of substrate 1d were undisturbed during the transformation; aziridine 2d was obtained in a 69% yield. Highly strained fused-ring and spirocyclic substrates gave corresponding products 2e and 2f, respectively, in 42% and 45% yields. Next, we subjected multisubstituted styrenes to electrolysis with NH3 in MeOH. The α-cyclopropylstyrene analogue 1g gave rise to 2g in a 52% yield. Several ethers were examined and found to afford products 2h– 2m. The yields of esters 2n and 2o were low (only 43% and 34%), which was attributed to partial decomposition of the ester groups by NH3. Product 2p, which contains a sulfamate group, was prepared in 53% yield. The yields of 2q and 2r, which have heterocyclic substituents with multiple nitrogen atoms, decreased during purification (41% and 48%). We determined that the conjugation in substrates 1s and 1t had no effect on the reaction: the corresponding aziridines with a terminal vinyl group ( 2s) and a terminal ethynyl group ( 2t) were obtained in 53% and 51% yields, respectively. The structure of tetraphenyl aziridine 2u was confirmed by X-ray analysis of its crystal. Figure 2 | Substrate scope of electrochemical aziridination using NH3. aConditons: GF anode and Pt cathode, 1 (0.2 mmol), Mg(ClO4)2 (0.1 mmol), MeOH (5.0 mL), 6.5 V or 7.0 V cell potential, 0 °C, 4 h. bGF anode and Pt cathode, Mg(ClO4)2, MeOH/dichloromethane (DCM) = 3:1 (v/v), 6.5 V, 0 °C, 4 h. cGF anode and GF cathode, Mg(ClO4)2, MeOH/DCM = 4:3 (v/v), 7.5 V, r.t., 6 h. dGF anode and GF cathode, LiClO4, MeOH, 6.5 V, r.t., 4 h. eGF anode and Pt cathode, LiClO4, MeOH, 7.0 V cell potential, rt, 4 h. fLiClO4, EtOH, constant current, 0 °C, 6 h. Download figure Download PowerPoint The effects of steric bulk and conjugation pose challenges for the activation of the C=C bonds of triarylsubstituted ethylene, so we evaluated the utility of our electrochemical protocol for direct aziridination of these compounds. We began by elucidating the effects of various para substituents and found that moderate to good yields of 2v– 2ae could be obtained and 2af– 2ao could also be generated. Amide-containing substrates, including amides bearing allyl and propargyl groups, were compatible with reaction conditions, affording 2ap– 2aw. Finally, we were also able to synthesize diaryl-substituted aziridines 2ax– 2bc from the corresponding stilbenes in yields of around 50%. Mechanism investigation To understand the origin of the chemoselectivity, we chose triphenylethene ( 1v) as a model substrate because it showed the greatest contrast of chemoselectivity for aziridination (MeOH) versus hydrogenation (DMF; Supporting Information Table S2). Cyclic voltammetry (CV) in DMF showed that anodic oxidation of NH3 predominated (Figure 3a; Supporting Information Section 5.1.5), whereas in MeOH, 1v was oxidized prior to NH3 ( Supporting Information Section 5.1.4). This finding suggests that a Lewis acid–base interaction between MeOH and NH3 stabilized the lone pair electrons on nitrogen. To gain a computational understanding of the dramatic effect of the solvent on the chemoselectivity, we used a multistep explicit solvation protocol77–81 to calculate the ionization energy during single-electron oxidation (Eox) of NH3 and 1v in the DMF and MeOH, as well as in the gas phase (Figure 3b; Supporting Information Section 6.2, Figures S4–S6, and Tables S5 and S6). Our multistep explicit solvation protocol showed that in the gas phase, NH3 was much more difficult to oxidize than 1v. The presence of solvent DMF caused a small decrease in the Eox of 1v and a larger drop in the Eox of NH3. As a result, Eox(NH3) was only slightly (0.16 eV) higher than Eox( 1v) in DMF, suggesting interfering NH3 oxidation pathways. Changing the solvent to MeOH dramatically affected the Eox values. In MeOH, Eox(NH3) was 0.78 eV higher than Eox( 1v), a difference that decreased the probability of NH3 oxidation. Further analysis showed that this increase in the difference between the Eox values resulted from the combination of an uphill shift of Eox(NH3) and a downhill shift of Eox( 1v). We contend that ΔEox(NH3), which can be attributed to changes in hydrogen-bonding interactions between the solute and solvent, was the major contributor to the difference. The Kamlet–Taft solvation parameters82 of DMF and MeOH (β = 0.69 and 0.66, respectively) indicate that they have similar abilities to act as hydrogen bond acceptors. In contrast, their abilities to act as hydrogen bond donors were very different (α = 0 for DMF and α = 0.98 for MeOH), with MeOH being a strong donor. Therefore, we reasoned that NH3 oxidation accompanied a decrease in its ability to act as a hydrogen bond acceptor and thus that the hydrogen-bond-donating solvent MeOH was better at stabilizing the reactant and destabilizing the product. Thus, Eox(NH3) could be expected to be higher in MeOH. Figure 3 | (a–d) CV analysis and DFT computational study on the reaction pathway. aGlassy carbon working electrode and Pt counter electrode, 50 mV/s, 1v (0.01 mol/L), NH3 (0.02 mol/L) in DMF (LiClO4, 0.1 mol/L), in MeOH (LiClO4, 0.1 mol/L). bE1 is obtained from the multistep explicit solvation protocol calculations; E2 is derived by adding the DFT-computed E2−E1 value to the E1 value. Download figure Download PowerPoint Based on the above-described findings, we explain the anodic behavior of NH3 by combining the above-calculated Eox and further density functional theory (DFT) calculations (at B3LYP-D3(BJ)/6-311+G(d,p)//B3LYP-D3(BJ)/6-31+G(d) level of theory) on the solution-phase reaction pathways ( Supporting Information Section 6.1, Figures S1–S3, and Table S4). In the aprotic solvent DMF (Figure 3c), electron transfer between 1v radical cation and NH3 occurs readily with the increase of free energy of only 3.6 kcal mol−1. Subsequent deprotonation of NH3 radical cation is barrierless and leads to the production of hydrazine83 and, eventually, N2; whereas nucleophilic addition of NH3 to 1v radical cation has an activation barrier of 8.5 kcal mol−1. In contrast, in the protonic solvent MeOH, the formation of NH3 radical cation is endergonic by 17.9 kcal mol−1, much higher than the barrier for the NH3 addition to 1v radical cation via transition state TS1-a (8.7 kcal mol−1), which will lead to the aziridination of 1v. We propose that the reaction in MeOH proceeds via the pathway outlined in Figure 3d. The reaction starts anodically with the oxidization of 1v to its radical cation (Eox = 6.64 eV). NH3 then nucleophilically adds to the less-substituted site of 1v radical cation in a process with a free-energy barrier of 8.7 kcal mol−1. The resulting intermediate ( Int2-a) undergoes nearly barrierless deprotonation. Then a second oxidation, the calculated ionization energy 5.72 eV of which is lower than that of the initial oxidation of 1v, takes place. Subsequent ring closure leads to the formation of a protonated aziridine moiety, which undergoes deprotonation to afford the product. The generated NH4+ cations are reduced at the cathode to evolve H2. So, the DFT computational study shows that MeOH ensures the desired chemoselectivity toward aziridination via a two-electron oxidation instead of hydrogenation of alkene 1v. Furthermore, the substantial difference between the Pt and GF anodes when MeOH (Table 1, entry 5) was applied could be explained. With CV analysis, we found that the oxidation of MeOH on the Pt anode proceeded much more readily than the oxidation of 1v ( Supporting Information Section 5.1.3).84 GF has a large carbon surface ( Supporting Information Section 5.2), which facilitates its affinity and oxidation for alkenes. Further application of aziridination using NH3 Finally, we explored the synthetic applications of our aziridination protocol (Figure 4). We began by carrying out a reaction of 1a on a 10-g scale using GF for both electrodes and LiCl instead of Mg(ClO4)2 as the supporting electrolyte (Figure 4a; Supporting Information Section 2.4). Under these conditions, aziridine 2a was obtained in a 65% isolated yield. Since the transformations of tetrasubstituted aziridines had not been extensively explored, we subjected 2a to a variety of reaction conditions ( Supporting Information Section 11). We found that the three-membered ring could be opened by Pd-catalyzed hydrogenation to afford α-tertiary amine 4 in 46% yield. In addition, 2a underwent acid-catalyzed ring-opening reactions with various nucleophiles to afford difunctionalized products 5– 8. Heating 2a in trifluoroacetic acid resulted in an intramolecular ring-opening/annulation cascade that afforded indole derivative 9 in 58% yield. However, when 2a was treated with a basic nucleophile, such as MeLi or EtMgBr, no reaction occurred, even at elevated temperature; the starting material was recovered unchanged (data not shown). However, the anisole group moiety could be modified by such nucleophiles; for example, we accomplished C–H silylation to afford 10 and Ar-OMe phenylation to afford 11. In addition, we found the tetraphenyl aziridine 2u exhibited fluorescence in the mixture of tetrahydrofuran (THF) and water and in solid state ( Supporting Information Section 12), demonstrating the constrained conformation of 2u could lead to aggregation-induced emission (AIE) effect even without conjugation alkene moiety (Figure 4b).85 Figure 4 | (a and b) Further utilization of azirdines in chemical transformation and as AIE material. Download figure Download PowerPoint Conclusion We have demonstrated that NH3 can be used as the ultimate nitrogen source in electrochemical aziridination reactions. Specifically, we found that NH3 directly reacted with tetrasubstituted alkenes to afford tetrasubstituted aziridines, which are difficult to access by other methods. The use of GF as the anode material and MeOH as the solvent was critical for achieving the desired aziridines, and the reasons for these findings were elucidated by means of CV and DFT calculations. The reaction can be easily scaled up, and one of the products was used as a precursor for the synthesis of a variety of bulky amines. Our findings indicate that the use of NH3 as a nitrogen source in electrochemical syntheses merits further exploration. Supporting Information Supporting Information is available and includes the general procedures for electrochemical aziridination and electrochemical hydrogenation, additional optimization results, electrochemistry analyses, computational details, crystal structures, characteristic data, and spectra of new compounds. Conflict of Interest The authors declare no competing interest. Acknowledgments The authors are grateful to the High-Performance Computing Center of Nanjing University, Collaborative Innovation Center of Advanced Microstructures, and the National Supercomputing Center in Wuxi for performing the calculations at their computing facilities. This work was supported by the National Natural Science Foundation of China (nos. 22071105, 22031008, 21803030, and 22001089), QingLan Project of Jiangsu Education Department, and the Jiangsu Innovation & Entrepreneurship Talents Plan in China. J.L. appreciates the support from the Nature Science Foundation of Jiangsu Province (no. BK20191046).