Rational design of *N*-heterocyclic compound classes via regenerative cyclization of diamines

The discovery of new chemical reactions is a crucial area of research, particularly when it enables access to previously unsynthesized compound classes. *N*-heterocyclic compounds are of significant importance due to their widespread applications in materials science and life sciences (pharmaceuticals, agrochemicals, dyes, conductive materials). This study focuses on developing a rational design strategy for synthesizing these compounds. Traditional approaches often lack this rational design element. Iterative synthesis, where a modified functional group is regenerated, offers a route to introduce chemical diversity and potentially address functional or global challenges. Metal-catalyzed reactions have emerged as powerful tools for automated C–C bond formation and selective olefin syntheses. This research explores the concept of regenerative cyclization, where a ring closure reaction regenerates the original set of functional groups, as a means to synthesize cyclic compounds. Specifically, the focus is on using amines as the key functional groups to access classes of *N*-heterocyclic compounds. A catalytic consecutive three-component reaction is developed, initiating with a diamine, an amino alcohol (undergoing dehydrogenation, condensation, and cyclization to generate a new pair of amines), and a subsequent ring closure with an aldehyde, carbonyldiimidazole, or a dehydrogenated amino alcohol. The Earth-abundant metal manganese is employed as a catalyst for the dehydrogenation step. This method exhibits diastereoselectivity, broad functional group tolerance (even hydrogenation-sensitive groups), and scalability. Notably, all synthesized *N*-heterocyclic compounds are novel.

The synthesis of 2,3-dihydro-1*H*-perimidines from 1,8-diaminonaphthalene and aldehydes is a known reaction (reported in 1964). Recent work has demonstrated the catalytic generation of aldehydes for such couplings via dehydrogenation catalysis using phosphine-free manganese complexes. The authors cite previous research on iterative synthesis and regenerative cyclization, highlighting the potential of these approaches to create chemical diversity and access new compound classes. Several relevant studies focusing on metal-catalyzed reactions for automated C–C bond formation and selective olefin syntheses are reviewed, showcasing the growing interest in utilizing metal catalysis for the rational design of chemical synthesis. The importance of *N*-heterocyclic compounds in various applications is also discussed, providing the context for the study's objective. Existing manganese-catalyzed hydrogenation and dehydrogenation reactions, along with the use of pincer ligands, are cited to support the selection of the catalytic system employed in this research. Specifically, the authors reference studies using PN^5P-pincer ligands and their effectiveness in manganese-catalyzed dehydrogenative coupling of alcohols and amines.

The research began with reaction optimization for the synthesis of 2-(2,3-dihydro-1*H*-perimidin-2-yl)aniline (A1) from 1,8-diaminonaphthalene and 2-aminobenzyl alcohol. Various earth-abundant metal complexes (Mn, Fe, Co) stabilized by pincer ligands were screened as precatalysts. Manganese catalysts with a PN^5P-pincer ligand exhibited the highest activity. Reaction parameters such as temperature, catalyst loading, solvent, and base were optimized (details provided in Supplementary Tables 1–7). The optimal conditions for A1 synthesis involved 1 mol% [Mn-I] precatalyst, 30 mol% KO*t*Bu, 3 mL 2-MeTHF, and 100 °C for 2 h in an open system. The substrate scope was investigated using 21 aminobenzyl alcohol derivatives, yielding the corresponding 2,3-dihydro-1*H*-perimidines (A1–A21; 'amino perimidines'). The reaction tolerated electron-donating and -withdrawing groups, as well as various functional groups such as methoxy, dimethoxy, trifluoromethoxy, and acetal groups. Single-crystal X-ray diffraction confirmed the structure of A1. The primary amine functionality in the amino perimidines was used for a second ring closure (modification degree 2) using aldehydes. This led to a new class of compounds, termed 'fertigines,' which consist of two six-membered *N*-heterocyclic ring systems. The fertigines were synthesized via a consecutive multi-component one-pot reaction, starting with the optimized conditions for amino perimidine synthesis, followed by aldehyde addition. The substrate scope of fertigines was explored with various aldehydes (including halogenated, methylated, methoxy-substituted, heterocyclic, and aliphatic aldehydes), demonstrating the breadth of this reaction. The influence of substituted 1,8-diaminonaphthalenes on the synthesis of amino perimidines and fertigines was also investigated. Upscaling experiments were performed, confirming similar yields in multigram-scale synthesis. The synthesis of amino alkyl perimidines from aliphatic amino alcohols was investigated, followed by a ring closure reaction with N,N'-carbonyldiimidazole (CDI) to produce 'kuenstlerines'. Finally, the synthesis of amino fertigines via a second ring closure step with 2-aminobenzyl alcohols was explored. Mechanistic studies involving time-dependent 1H NMR and the isolation of an intermediate (A1K) were conducted to propose a mechanism for the catalytic cycle and ring closure cascade.

The study successfully developed a novel three-component reaction for the synthesis of various *N*-heterocyclic compounds, including a new class called fertigines, via regenerative cyclization. The reaction utilizes a manganese-based catalyst for amino alcohol dehydrogenation and exhibits high diastereoselectivity. A wide range of functional groups were tolerated in both the amino perimidines and fertigines synthesis. The reaction demonstrated a broad substrate scope, accommodating various aminobenzyl alcohols and aldehydes. The use of substituted 1,8-diaminonaphthalenes also proved successful, generating a diverse set of products. The synthesis of amino alkyl perimidines and their subsequent conversion to kuenstlerines using CDI was also achieved. The reaction was easily scaled up, maintaining similar yields in multigram synthesis. Mechanistic studies suggested a base-mediated cyclization and provided evidence for a proposed mechanism involving an intermediate (A1K). The yields obtained for amino perimidines (A1-A24) ranged from 69% to 97%, while fertigines (B1a-B5c) yields ranged from 56% to 95%. Amino alkyl perimidines (A25-A27) yields were 91-94%, and kuenstlerines (C1-C3) yields were 76-91%. Amino fertigines (B6a-B6c) were synthesized in 38-79% yield. All synthesized compounds are novel, lacking CAS numbers.

The results demonstrate the effectiveness of regenerative cyclization as a strategy for the rational design of *N*-heterocyclic compounds. The use of manganese catalysis in combination with a three-component reaction allows for the efficient and selective synthesis of a wide array of structures. The broad substrate scope and functional group tolerance expand the synthetic possibilities for these important molecules. The discovery of the novel fertigine and kuenstlerine classes significantly contributes to the field of *N*-heterocyclic chemistry. The mechanistic insights gained provide a foundation for further optimization and development of similar reactions. The scalability of the reaction highlights its potential for practical applications. The successful use of various building blocks, including aliphatic and heterocyclic derivatives, underscores the versatility and robustness of this new method. The methodology presented offers a sustainable and efficient way to synthesize diverse *N*-heterocyclic compounds that are difficult or impossible to obtain via traditional methods.

This research successfully demonstrated the synthesis of novel *N*-heterocyclic compounds using a regenerative cyclization approach. The manganese-catalyzed three-component reaction exhibited excellent substrate scope, functional group tolerance, and scalability. The discovery of two new classes of compounds, fertigines and kuenstlerines, significantly expands the library of available *N*-heterocycles. Future research directions could focus on exploring other diamines and ring-closing reagents to further broaden the range of accessible compounds. Investigating the biological activities of the synthesized compounds would also be a valuable area of future study. Optimizing the reaction conditions for even higher yields and exploring the use of alternative catalysts is another avenue for future research.

While the reaction demonstrated a broad scope, certain substrates led to lower yields, possibly due to steric effects or electronic properties. The synthesis of amino fertigines yielded lower yields than the other products synthesized which warrants further investigation. Some of the products showed limited air stability, requiring careful handling and storage. The mechanistic studies are based on experimental observations and require further investigation to fully elucidate the catalytic cycle. Although the methodology shows scalability, further studies may be needed to fully optimize the process for large-scale industrial applications.