Visible-Light-Induced Radical Cascade [4 + 2]/[4 + 2] Cycloaddition of Underexplored N‑Acryloyl Indoles To Access Dihydropyrido[1,2‑a]‑indolones

Abstract

We report a visible-light-mediated radical cascade [4 + 2]/[4 + 2] cycloaddition of simple N-acryloyl indoles and N-hydroxyphthalimide esters, which provides a streamlined route to structurally complex dihydropyrido­[1,2-a]-indolones. The reaction features a sequence of four consecutive radical additions involving two N-acryloyl indole molecules, which forges four new C–C bonds and induces dearomatization of one indole ring. The proposed mechanism is supported by observations of [4 + 2] cycloaddition byproducts and additional control experiments. The synthetic utility of this method is demonstrated by the scale-up reaction and downstream derivatizations.

Nitrogen-fused polycyclic indoles, such as dihydropyrido­[1,2-a]-indolones (DHPIs), are widespread structural motifs in natural products and pharmaceutical compounds, many of which display notable pharmacological properties and biological activities (Figure ). Consequently, substantial efforts have been devoted to their construction by both synthetic and medicinal chemists, leading to the development of numerous elegant methods. Among these, radical cascade cyclization has distinguished itself as one of the most powerful and well-studied strategies, due to its ability to streamline synthetic sequences, reduce costs, and minimize waste generation. To access N-fused polycyclic indoles via radical cascade cyclization, two general approaches are typically employed: , (i) reactions of N-alkene/alkyne-tethered indoles with external radical precursors and (ii) couplings of indoles bearing N-tethered radical precursors, either through self-coupling or with external alkenes. Both approaches have been successfully realized under various reaction conditions, including photocatalysis, electrocatalysis, and oxidant-promoted processes, all of which proceed through a fundamentally similar mechanism. , Typically, an initial single-electron transfer (SET) event generates a reactive radical species from a radical precursor. This radical intermediate then undergoes a series of inter- and/or intramolecular additions to π bonds, driving the cascade cyclization. A final SET step followed by deprotonation ultimately furnishes the desired N-fused polycyclic indole products.

1.

Representative bioactive DHPIs.

Specifically, 2-aryl-N-acryloyl indoles and 1-acryloyl-2-cyanoindoles are particularly compelling members of the N-alkene-tethered indole family, , owing to their dual reactive sites that serve as excellent radical acceptors for radical cascade cyclizations. 2-Aryl-N-acryloyl indoles readily engage with a wide range of carbon-, nitrogen-, sulfur-, silicon-, phosphorus-, and germanium-centered radical precursors through transition-metal (TM)-driven, oxidant-mediated, visible-light-induced, or electro-promoted radical cascade cyclizations, enabling efficient access to a large variety of indolo­[2,1-a]­isoquinolines. ,,− Similarly, pyrrolo­[1,2-a]­indolediones can be obtained by treating 1-acryloyl-2-cyanoindoles with different radical precursors in the presence of stoichiometric persulfate under blue light irradiation. Although impressive progress has been achieved in the study of these 2-substituted-N-acryloyl indoles, their structural analoguessimple N-acryloyl indolesare still largely underexplored, despite being more easily synthesized and accessible (Scheme a). Only a handful of studies on their reactivity toward N-fused polycyclic indoles have been documented to date. The initial work was introduced by the Kerr group in 2008,^12a who assembled DHPIs by reacting simple N-acryloyl indoles with Michael donors in the presence of base and Mn­(OAc)₃. However, further advances did not appear for over a decade. In 2020, Paton, Smith, and co-workers disclosed a blue-light-mediated Ir­(III)-catalyzed [2 + 2] cycloaddition of simple N-acryloyl indoles, offering a facile route to pyrrolo­[1,2-a]­indol-3-ones. More recently, the Liu and Guo groups showed both photo- and electro-induced radical cascade [3 + 2]/[4 + 2] cyclizations by treating simple N-acryloyl indoles with 2-(iodomethyl)­cyclopropanes or α-allyl-activated methylenes, respectively. , These newly developed transformations are particularly intriguing because they forge three new C–C bonds through sequential radical additions, whereas previous work only generates one or two σ bonds. Despite these notable advances, the field remains underdeveloped, as reported examples are scarce to date. Additionally, these existing studies rely on uncommon radical precursors and are restricted to C2–functionalization of the indole cores. Furthermore, to the best of our knowledge, there are no prior literature reports describing the dimerization of 2-substituted or simple N-acryloyl indoles for the synthesis of N-fused polycyclic indoles. Taken together, the reactivity study of simple N-acryloyl indoles is still in its infancy. Therefore, exploring their additional reactivity (e.g., dearomatization or dimerization) with commonly used radical precursors is of considerable importance, as it would not only fill the current research gap but also overcome existing limitations, thereby opening new avenues for accessing structurally distinct N-fused polycyclic indoles (e.g., DHPIs).

1. a) Current Reactivity Study of Simple N-Acryloyl Indoles toward N-Fused Polycyclic Indoles and b) This Work.

In this work, we report a visible-light-promoted radical cascade [4 + 2]/[4 + 2] cycloaddition of simple N-acryloyl indoles (1) with N-hydroxyphthalimide (NHPI) esters, which provides a straightforward and efficient route to structurally complex DHPIs (Scheme b). The proposed mechanism is shown in Scheme . The excited state of the photocatalyst (PC*) reduces the radical precursor via oxidative quenching, generating the radical species I (R^•) and PC⁺. Radical I then triggers a cascade of three consecutive radical additions: two intermolecular additions to the alkene moieties of two simple N-acryloyl indole molecules, followed by an intramolecular addition to the C2C3 bond of the indole ring, producing the key benzyl radical intermediate IV. Intermediate IV can diverge into two possible pathways. In Path A, it engages in a fourth radical addition to form another benzyl radical species V, which subsequently undergoes SET coupled with deprotonation to yield the desired [4 + 2]/[4 + 2] product. In Path B, an alternative [4 + 2] cycloaddition byproduct may form via direct SET and deprotonation from intermediate IV. We envisioned that the intramolecular radical addition in Path A is favored over the competing SET in Path B due to its inherent spatial proximity, thereby selectively delivering the [4 + 2]/[4 + 2] adduct as the major product. Given the broad bioactivity of DHPIs (Figure ), developing efficient methods for their synthesis is highly desirable. The methodology described herein constitutes a major advance over existing approaches. On one hand, it accommodates two common types of radical precursors (e.g., NHPI esters and α-bromocarbonyl compounds) while enabling the formation of four new C–C bonds and inducing dearomatization of one indole ring. More importantly, the dimerization of N-acryloyl indole molecules is entailed during the formation of final productsan outcome not realized using 2-aryl-N-acryloyl indoles, 1-acryloyl-2-cyanoindoles, or simple N-acryloyl indoles in previous reports.

2. Proposed Reaction Mechanism.

As part of our ongoing investigation of redox-active NHPI esters in photocatalysis, simple N-acryloyl indole 1a and 1,3-dioxoisoindolin-2-yl isobutyrate 2a were selected as model substrates to optimize reaction conditions (Table ). Given the reduction potentials of alkyl NHPI esters (E _red = −1.20 to −1.37 V vs SCE in MeCN), we initiated the study using fac-Ir­(ppy)₃ (E _IV/_*III = −1.73 V vs SCE in MeCN) under purple LEDs (λ_max = 390 nm) in DMSO at room temperature under Ar for 18 h. To our delight, the proposed [4 + 2]/[4 + 2] product 3a was obtained in 60% isolated yield (entry 1). Solvent screening confirmed DMSO as optimal (entries 1–5), while alternative PCsincluding [Ru­(bpy)₃Cl₂]·6H₂O, Eosin Y, or tetraphenylporphyrin (TPP)did not improve yields (entries 6–8). Considering the deprotonation step in the proposed mechanism presented in Scheme , we evaluated the effect of external bases: NaHCO₃ led to a messy mixture, whereas 2,4,6-collidine afforded 3a in 57% yield (entries 9–10). Reducing PC loading from 3 to 1 mol % had minimal impact (entry 11). White light irradiation (23 W CFL) increased the yield to 62% and was adopted as the optimal condition (entry 12). Control experiments highlighted key features of the reaction: exposure to air led to messy outcomes (entry 13), no product formed in the dark (entry 14), and only trace 3a was observed in the absence of PC under 23 W CFL (entry 15). Notably, 3a was obtained in 49% yield under purple LEDs without PC (entry 16), likely due to the formation of photoactive electron donor–acceptor (EDA) complexes. More optimization studies are summarized in Tables S1–S2.

1. Selected Optimization of Reaction Conditions.

^aThe reaction was conducted with 1a (0.3 mmol), 2a (0.1 mmol) and PC (x mol %) in solvent (1.0 mL).

^bIsolated yield.

^cNMR yield by using trimethoxybenzene (0.1 mmol) as the internal standard.

^dOpen to air.

^eNo reaction.

With the optimized conditions in hand, we next evaluated the generality of the reaction (Table ). We first explored the scope of NHPI esters 2 using 1a as a model substrate (Table a). The reaction proceeded smoothly, affording DHPIs 3a–3p in 14–71% yield. A broad spectrum of functional groups is well-tolerated, including aryl (3h, 3o), 2-thienyl (3i), chloride (3j), ketone (3k), ester (3l), ether (3m–3n), and Boc-protected amine ester (3p). All NHPI ester substrates 2 produced a single diastereomer, except the chiral aspartic acid-derived ester 2p, which gave the corresponding DHPI 3p in 26% yield with a 1.3:1 d.r. value. The relative configurations of 3a and 3j were confirmed by X-ray crystallography (CCDC 2465195 and 2465294; see Supporting Information), showing the same stereochemical arrangement. Notably, several low-yield substrates exhibited drastic improvement under purple LEDs (λ_max = 390 nm): 3b (14 → 48%), 3l (43 → 54%), and 3p (26 → 51%). Next, we examined the scope of simple N-acryloyl indoles 1 utilizing 2e as a model substrate (Table b). Substituents (R¹ or R²) at C3–C7 positions reacted efficiently with 2e, furnishing an array of functionalized DHPIs 3q–3ae in 23–82% yield. Both electron-donating and electron-withdrawing groups (EWGs) were compatible. Nevertheless, products 3 with strong EWGs, such as CF₃ (3aa), CO₂Me (3ab), and F (3ad), were formed in slightly lower yields, presumably due to the less favored second SET process associated with formation of the less stable benzylic carbocation intermediate VI. To our delight, these yields also improved under purple light irradiation (3aa: 37 → 43%; 3ab: 23 → 49%; and 3ad: 44 → 48%). Interestingly, when substrate 1p bearing 5-methoxy and 7-methyl groups was tested, the desired product 3ae was isolated in 43% yield, along with the [4 + 2] byproduct 3ae′ in 10% yield (Table c). The isolation of 3ae′ further supports the mechanism proposed in Scheme , and its yield increased to 25% under purple light irradiation. It is worth noting that although [4 + 2] byproducts were observed for most substrates, their low formation levels and the presence of cis/trans isomers allowed isolation only of 3ae′. These observations help explain the generally moderate yields of 3. Finally, the protocol was ineffective for certain substrates (Table d). Product 3af was isolated but could not be characterized by NMR due to poor solubility, although HRMS confirmed its structure (Figure S1). Products 3ag–3ah and 3aj–3ak were not detected, and only trace 3ai was observed. The failure to form 3ag–3ah is presumably due either to direct quenching of fac-Ir­(ppy)₃ by the NO₂ group, as evidenced by the observation of unreacted 2e, or to an unfavored second SET process. Additional substrate scope investigations are summarized in Tables S3–S4 and Schemes S1–S3.

2. Scope of Simple N-Acryloyl Indoles and NHPI Esters .

^aReactions were conducted on a 0.1 mmol scale. Isolated yields.

^bPurple LEDs (λ_max = 390 nm) were used.

To showcase the synthetic utility of this approach, several additional experiments were performed (Scheme ). First, diethyl bromomalonate 4a and 2-bromoacetophenone 4b were investigated as radical precursors to react with 1a under the standard white light conditions, respectively. Unfortunately, both reactions provided messy results (Scheme S4). Given that deprotonation is required for product formation, 1.0 equiv of NaHCO₃ was added under the above optimal conditions. Delightfully, 5a and 5b were obtained smoothly, with yields increased to 50% and 41% under blue light (λ_max = 440 nm) irradiation (Scheme a). Next, the reaction was successfully scaled up to 4.2 mmol, albeit with slightly reduced efficiency, affording 3e in 46% yield (Scheme b). Finally, several derivatizations of product 3a were explored (Scheme c). Reduction of both amide groups with BH₃·Me₂S furnished 6a in 53% yield, whereas treatment with LiAlH₄ delivered the selectively reduced product 6b in 78% yield as a single diastereomer (see mechanism in Scheme S5). Although the relative configuration of 6b was not determined at this stage, it could be further converted to 6c via reaction with TsCl in the presence of triethylamine. Collectively, these results underscore the versatility, practicality, and robustness of this approach.

3. Synthetic Applications.

To gain further insight into the reaction mechanism, several control experiments were conducted. Radical trapping and light/dark interval experiments support the proposed mechanism in Scheme (Schemes S6–S7). Notably, the formation of 3a, 3b, 3p, and 3ab in the absence of PC under purple light suggests that an EDA-complex-driven pathway is also possible (Table S5). This alternative pathway could account for the increased yields observed for certain products (e.g., 3b, 3p, 3aa, and 3ab) indicated in Table . A detailed EDA-complex-driven mechanism is proposed in Scheme S8. The observed diastereoselectivity can also be rationalized by two proposed mechanistic models involving the key intermediate IV (Scheme S9).

In summary, we have developed a novel visible-light-mediated radical cascade [4 + 2]/[4 + 2] cycloaddition that uncovers previously unobserved reactivity of simple N-acryloyl indoles. This approach enables the efficient one-pot construction of structurally complex DHPIs with diverse functional groups, representing a significant advance over previously reported methods in three key aspects. First, it is compatible with two common classes of radical precursors: NHPI esters and α-bromocarbonyl compounds. Second, it forges four C–C bonds and induces dearomatization of one indole ring during the radical addition sequence. Third and most importantly, two N-acryloyl indole molecules are incorporated into the DHPI products. In addition, the observed [4 + 2] byproducts and mechanistic studies provide insight into the reaction pathways. The generally moderate yields are likely attributable to the formation of multiple reactive radical intermediates, which result in undetermined byproducts. The scale-up reaction and downstream transformations further highlight the synthetic utility of this methodology. Ongoing efforts aim to engage simple N-acryloyl indoles with other radical precursors under photocatalytic, electrocatalytic, or oxidant-induced conditions.

The data underlying this study are available in the published article and its Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.orglett.5c05421.

• Experimental procedures, characterization, X-ray crystallographic data, and NMR spectra (PDF)

†.

C.B. and S.M. contributed equally to this work.

The authors declare no competing financial interest.