Development of Annulative Coupling and Aromatic Amination toward the Synthesis of Structurally New Aromatic Amines

Abstract

Aromatic amines are widely recognized as some of the most important scaffolds in various functional molecules, such as pharmaceuticals and materials. Two reactions have been developed to access structurally novel aromatic amines: 1) annulative coupling and 2) aromatic amination. This account summarizes the recent progress in these two reactions.

1. Introduction

Aromatic amines are the most important class of compounds that are widely found in functional molecules, including biology-oriented aromatic amines^1,2 and materials-oriented aromatic amines^3,4 (Figure 1A). The diverse functions of these aromatic amines are attributed to the rigid structure of the aromatic core and presence of substituent amino groups. To access new aromatic amine molecules structurally, the author developed a general and divergent strategy by dividing the research into two distinct topics (Figure 1B). The first involves the construction of arene cores through annulative coupling reactions that allow the generation of new arene molecules. The second topic focuses on arene functionalization, particularly, aromatic amination, which enables the direct installation of an amine moiety into aromatic cores. By harnessing these two topics, the overall aim is to provide a new strategy to synthesize novel aromatic amines.

Representative functional aromatic amines and the strategy toward the synthesis of structurally new aromatic amines.

Note that the concept developed by the author is not entirely new but inspired by the works of his mentors, Yorimitsu, Oshima, Osuka, and Itami. Yorimitsu and Oshima studied the carbometallation chemistry of unsaturated hydrocarbons, explored by the author in his PhD study,^5,6 by converting alkynes into alkenylmetal intermediates, which can be regarded as the construction of a core structure. The resulting alkenyl metal exhibits versatile reactivity toward coupling reactions, thus enabling the synthesis of a variety of alkene compounds that can be regarded as functionalization of the core structure. Yorimitsu, Oshima, and Osuka reported the synthesis of multisubstituted benzofurans from phenols, where ketenedithioacetal monoxide (KDM) was reacted with phenols to yield 2-methylthiobenzofurans (core construction).^7,8 The sulfur group was subsequently converted to various functional groups through cross-coupling reactions (core functionalization). Although these two studies involved simple two-step syntheses, they provided a pathway to access a variety of functional molecules. Additionally, the author acknowledges the contributions of Itami to Chemistry. Itami established the programmed synthesis of tetraarylthiophenes, utilizing 3-methoxythiophene as the platform, in which the importance of developing elegant methodologies was highlighted.⁹ This simple but powerful strategy enables the synthesis of various thiophene derivatives. Moreover, when the author joined Itami’s group, Itami conducted interdisciplinary research by bridging organic chemistry with nanocarbon science^10–12 and plant biology,^13,14 which considerably influenced chemistry and the author’s way of thinking. The pursuit of new reactions and advancement of straightforward strategies are key to unlocking the possibilities of next-generation molecular science.

2. Annulative Coupling

In this chapter, annulative coupling reactions for the synthesis of new aromatic cores are described (Figure 2). Since Ullmann’s discovery of the reductive dimerization of aryl halides,¹⁵ this reaction has been widely employed in the synthesis of various biaryl-based functional molecules. In contrast to classical dimerizations, in which arenes are connected by a single bond, our annulative coupling enables the construction of new aromatic cores from simple monohalogenated arenes under Pd-catalyzed conditions.

Developed annulative coupling reactions.

The discovery of these annulative couplings dates back to 2015, when research was focused on pyridine functionalization chemistry,^16,17 and a new bromobenzene trimerization reaction was discovered (Figure 3A). When bromobenzene is subjected to the Pd-catalyzed conditions, trimerization occurs and yields triphenylene. Although numerous organic chemists have studied palladium chemistry using haloarenes, this reaction remained unreported until the discovery of its transformation. However, this trimerization was found to be less synthetically useful because of the formation of isomers upon reaction with a substituted bromobenzene (Figure 3B). For example, the reaction of 4-bromotoluene affords a mixture of two trimethyltriphenylene isomers. Consequently, we explored this reaction using bromobiphenyl as a substrate. Surprisingly, instead of trimerization, a new dimerization reaction occurred and yielded phenyl-substituted triphenylene (Figure 3C). This reaction can be regarded as a novel type of dimerization of halobiphenyls that proceeds via annulation.^18,19 As mentioned above, our annulative dimerization is conceptually novel, as the reaction proceeds in a redox-neutral manner with the generation of HX (Figure 4).

Serendipitous findings.

Concept of annulative dimerization.

After investigating the reaction conditions, we successfully determined the optimal conditions for the synthesis of dimer 2a from chlorophenylene 1a. The treatment of chlorophenylene 1a with cesium carbonate in the presence of a palladium catalyst in cyclopentyl methyl ether (CPME) at 140 °C furnished dimer 2a in 81% yield (Figure 5). This reaction was applicable to various substituted phenylenes. For example, t-butyl-substituted phenylene afforded 2b in 77% yield. Furthermore, various other substituents, such as methoxy, trifluoromethoxy, methylthio, trifluoromethyl, trimethylsilyl, methoxycarbonyl, and carbazolyl groups, were compatible with the reaction conditions to yield 2c–2i in moderate to good yields. m-Methyl- and m-phenyl-substituted phenylenes smoothly underwent the reaction to afford the corresponding dimers 2j and 2k in 85% and 58% yields, respectively. Chlorobenzene derivatives with naphthyl or benzothienyl groups were applicable to the reaction to yield dimer 2l–2n in moderate to good yields. Further, 3-chloro-2,5-diphenylthiophene was dimerized to yield 2o in 33% yield. Note that not only chlorophenylenes but also phenylene triflates were dimerized under similar reaction conditions (Figure 6).²⁰ Relatedly, Shi reported the annulative iodophenylene dimerization as a useful route for polycyclic aromatic hydrocarbons (PAHs).²¹

Scope of annulative dimerization. *2h/2h′ = 3:1. Compound 2h′ is an isomer of 2h. ^†CsF is used instead of Cs₂CO₃.

Annulative dimerization of phenylene triflates.

Mechanistic experiments were performed to elucidate the reaction mechanism. The reaction with 1b afforded a mixture of 2p and 2p′ in 72% yield (Figure 7). Figure 8 shows the three potential reaction pathways. In the first step of the reaction, Pd(0) reacts with 1b to give intermediate A. Intermediate A can generate Pd-aryne intermediate B, which then reacts with 1b to produce the mixture of 2p and 2p′ (path (i)). Another possibility is the formation of palladacycle C from intermediate A through intramolecular C–H palladation. Intermediate C can yield both 2p and 2p′ (path (ii)). Although ortho-C–H palladation to yield intermediate D should be considered because of the increased acidity of the ortho-proton of the chloride atom, this can only result in the formation of 2p, which contradicts the experimental results; thus, this possibility should be excluded. As clarifying the reaction mechanism through only experiments is challenging, we performed calculations as well.¹⁹ Based on the computations, the reaction is likely to proceed through path (ii). A plausible reaction mechanism is shown in Figure 9. Intermediate C undergoes oxidative addition with 1b to yield intermediate E. From E, reductive elimination occurs to yield intermediates F and F′ as isomers. Product 2p is furnished from intermediate F through intramolecular C–H palladation and reductive elimination. Similarly, isomer 2p′ is generated from intermediate F′.

Mechanistic experiment. ^‡2p/2p′ = 3:1.

Possible reaction intermediates.

Plausible reaction mechanism based on calculation.

Dimerization can be used for further transformations (Figure 10). Dimer 2a was converted into 3, a fully fused PAH, in 77% yield. Interestingly, the reaction of single-bond-connected dimer 4 failed to yield 3, which highlights the advantage of annulative dimerization for synthesizing fully fused PAHs. This approach could be extended to the synthesis of larger PAHs (Figure 11). The treatment of 1c in our annulative dimerization conditions afforded 2q in 64% yield, which was subsequently converted to 5 in 72% yield.

Further transformation.

Short-step synthesis of fully fused PAH.

We extended our annulative coupling to bay-chlorinated phenanthrene derivatives (Figure 12).^22,23 The reaction of 6a in the presence of a palladium catalyst afforded the octagon-embedded 7a in 89% yield. Interestingly, although the reaction conditions were similar to the 6-membered formation conditions, only 8-membered formation occurred. The reaction was also applied to chlorotriphenylene, thereby affording 7b in 74% yield. Benzothiophene-substituted chlorophenanthrene dimerized to afford 7c in 84% yield. The resulting dimer 7a was converted to dibenzochrycene dimer 8 in 81% yield via an iron-mediated dehydrocyclization reaction. Owing to the lower solubility of chlorodibenzochrycene, accessing compound 8 through only dimerization is challenging.

Scope of octagon-forming annulative dimerization.

Although the reaction mechanism has not been fully elucidated, we propose the following reaction mechanism (Figure 13). Starting material 6b reacts with Pd(0) to yield palladacycle G through oxidative addition and intramolecular C–H palladation. Subsequently, palladacycle G reacts with 6b to generate intermediate H, which undergoes intramolecular C–H palladation to form spiropalladacycle I. Finally, a twofold reductive elimination occurs to yield product 7b through intermediate J.

Possible reaction mechanism of chlorophenanthrene dimerization.

We investigated coupling partners for cross-selective annulative coupling. In the initial trial of cross-selective annulative coupling, we employed chloro-, tosyloxy-, and bromobiaryls (Figure 14). However, no cross-coupling products were observed. Biphenylene was identified as the most effective partner for highly cross-selective reactions (Figure 15). A 1:1 mixture of chlorophenanthrene 6b with biphenylene (9) afforded the corresponding cross-coupling product 10 in 78% yield. Although the reaction conditions could potentially dimerize 6b or 9, dimers of these starting materials were barely detected. Owing to the cross-selective conditions, a three-fold annulative cross-coupling reaction was performed (Figure 16). The reaction of 11 with a slight excess of biphenylene (9) afforded 12 in 76% yield. The structure of major diastereomer 12a was unambiguously confirmed by X-ray crystallographic analysis.

Trial for octagon-forming annulative cross-coupling.

Discovery of octagon-forming annulative cross-coupling.

Three-fold annulative cross-coupling.

3. Aromatic Amination

The starting point of aromatic functionalization research was the Cu-catalyzed imidation of aromatics. To access new aromatic amines, a straightforward amination reaction of aromatic cores must be developed.^24,25 Notably, Ritter²⁶ and Baran²⁷ previously reported aromatic amination using arenes as the limiting reagent, which was accomplished using nitrogen radicals. We are interested in the nitrogen radical, which can be generated from a copper catalyst with N-fluorobenzenesulfonimide (NFSI).²⁸ Treatment of an arene with NFSI in the presence of CuBr and 6,6′-dimethylbipyridyl (6,6′-Me₂bpy) afforded the corresponding imidated products (Figure 17).^29,30 The scope of this reaction is sufficiently broad that various arenes such as PAHs, heteroarenes, natural products, and porphyrins can be imidated. For example, fluoranthene afforded 13a in 77% yield. Similarly, pyrenes were regioselectively imidated to produce 13b and 13c in 45% and 63% yield, respectively. Reactions with the thiophene derivative produced 14a and 14b without the loss of the bromo or acetyl groups. Interestingly, the reaction of 3-phenylthiophene produced 14c exclusively, and the imidation occurred at the hindered 2-position. When caffeine and flavones were subjected to these reaction conditions, the corresponding imidated products 15a and 15b were obtained in 61% and 78% yields, respectively. The reaction of the porphyrin proceeded at the meso-position to yield 16a. In the case of meso-tetraarylporphyrin, the β-position was imidated under the reaction conditions.

Representative examples of copper-catalyzed aromatic amination.

In this reaction, the ligand effect is significant (Figure 18). The use of 6,6′-Me₂bpy resulted in the highest yield for the transformation of fluoranthene to 13a. However, the yield decreases when 6-Mebpy is employed in the reaction. The presence of a methyl group adjacent to the nitrogen is critical as the yields significantly decreases with 5,5′-Me₂bpy, 4,4′-Me₂bpy, or bpy.

Ligand effect.

KIE experiments were performed to investigate the reaction mechanism. Interestingly, the intermolecular KIE value was determined to be 0.91 (Figure 19). Additionally, we demonstrated that electron-rich arenes react preferentially under these reaction conditions (Figure 20). In the competitive experiment, a mixture of 2-phenylpyridine and trifluoromethylated 2-phenylpyridine was treated with NFSI. Consequently, the more electron-rich 2-phenylpyridine preferentially reacted to yield 14e as the major product. These results support our working hypothesis that the imidyl radical is the key intermediate.

KIE experiment.

Competitive experiment.

To clarify the reaction mechanism, we performed calculations. An overview of this reaction is shown in Figure 21. Initially, the copper catalyst K reacts with NFSI to form copper dimer L. Dimer L is oxidized with NFSI, thus simultaneously generating the imidyl radical to yield intermediate M. Single electron transfer (SET) occurs from M to 2-phenylthiophene to yield complex N; from N, an addition reaction occurs to form radical O. A second SET forms P, which finally undergoes deprotonation and rearomatization to afford 14e. Based on the aforementioned reaction mechanism, we hypothesized that the imidation occurs at the most electrophilic site of the arene radical cation (Figure 22a). To validate this hypothesis, the Hirshfeld charges of the radical cations were calculated. Computational analysis using fluoranthene demonstrates that the position indicated by the red circle exhibits the highest Hirshfeld charge, which corresponds to the site of imidation. This hypothesis also applies to various arenes (Figure 22a). These calculations were performed on substrates for which the experimental results had already been obtained. To predict the reaction sites, we first conducted calculations on the two substrates (Figure 22b). Subsequently, we performed an imidation reaction on these substrates, which proceeded regioselectively to yield 14g and 14h as exclusive products at the predicted positions.

Calculated reaction mechanism.

Prediction of regioselectivity.

We also reported Au-catalyzed aromatic imidation with NFSI (Figure 23).³¹ Compared with the Cu-catalyzed conditions, the scope of this Au-catalyzed reaction is rather limited; however, various PAHs can be imidated to yield the corresponding products 13d–13f in moderate yields.

Gold-catalyzed aromatic imidation of PAHs.

Although the scope of Cu-catalyzed reactions is broad, nitrogen sources are limited. As NFSI is typically utilized as a fluorinating reagent, its derivatives are not commercially available. Thus, an additional step to fluorinate sulfonimides with F₂ is required. To circumvent this fluorination step, a novel C–H/N–H-type oxidative aromatic amination was developed. After this investigation, we determined the conditions using Ru-photoredox catalysis (Figure 24).^32,33 The treatment of arenes with sulfonimides in the presence of [Ru(bpy)₃]Cl₂•6H₂O as the photoredox catalyst and IBB (hypervalent iodine; the structure is shown in Figure 25) as the oxidant under blue light irradiation produced the corresponding imidated products. Owing to the oxidative conditions, various sulfonimides could be employed (17a–17c). The scope of this reaction is as broad as that of the Cu-catalyzed reaction. Various imidated PAHs (18a–18c) and heteroarenes (19a–19c) were obtained via this reaction.

Representative examples of photoredox-catalyzed aromatic amination.

Proposed reaction mechanism.

The proposed reaction mechanism for the photoredox-catalyzed imidation is shown in Figure 25. First, Ru(II) is excited by blue-light irradiation. The resulting Ru(II)* is oxidatively quenched with IBB to form Ru(III) and radical T. Ru(III) oxidizes sulfonimide to yield imidyl radical R, which subsequently reacts with naphthalene to form S. Single electron transfer from radical S to T provide cationic intermediate V and benzoate U. Finally, V aromatizes to yield the corresponding product.

For general aromatic imidation reactions, we constructed a chemical collection of synthesized aromatic amines. We identified an imidated oxazole, named SIM1, as a new class of bioactive molecules that inhibits stomatal opening (Figure 26).³⁴ Subsequently, we synthesized various SIM1 derivatives using the procedure shown in Figure 27 to construct a second set of chemical libraries (Figure 28). Through this process, we identified SIM3* as having higher activity and increased specificity for stomatal regulation. This study demonstrated the importance of developing new reactions to discover novel bioactive molecules.

Aryl sulfonimide SIM1 inhibiting stomatal opening.

Synthetic procedure of SIM1 derivative.

Second set of SIM1 derivative library.

4. Conclusion

Herein, the author summarized the developments made in the two reactions of annulative coupling and aromatic amination. By combining these two reactions, various aromatic amines can be synthesized. Furthermore, we identified a new bioactive molecule from a collection of aromatic amines. Currently, our group at the Kwansei Gakuin University is investigating alkylamines and ammonium salts, which may serve as new candidates for potential bioactive molecules.

Acknowledgment

The author is grateful to Prof. Kenichiro Itami for his insightful advice and guidance. The author wishes to express great appreciation for all the co-authors and collaborators. The author appreciates financial supports from JSPS KAKENHI Grant Numbers JP26888007, JP15K17821, JP17H04868, JP19K22251, and JP19H02700, Japan Science and Technology Agency (JST) PRESTO Grant Number JPMJPR20D8, the Sumitomo Foundation, the Uehara Memorial Foundation, Shorai Foundation for Science and Technology, the Naito Foundation, CASIO Science Promotion Foundation, TOBE MAKI Scholarship Foundation, the Noguchi Institute, Foundation Advanced Technology Institute, and Toshiaki Ogasawara Memorial Foundation.

References

Kei Murakami

Kei Murakami was born in Osaka in 1985 and grew up in Sapporo. He received his Ph.D. Degree from Kyoto University in 2012. After his JSPS postdoctoral fellowship, he started his academic career at the Hakubi Center for Advanced Research/Department of Chemistry, Graduate School of Science, Kyoto University. He then moved to Nagoya University in 2014 as Assistant Professor and became Designated Associate Professor in 2016. Since 2020, he has been Associate Professor at Kwansei Gakuin University, Japan.