3-Alkyl-3-hydroxy-2-oxindoles represent a privileged structural motif widely found in numerous biologically active natural products and synthetic pharmaceuticals,^[1] such as Convolutamydine A, Donaxaridine, Dioxibrassinine, Maremycin A, and Arundaphine (Figure 1).^[2] Consequently, their synthesis has attracted considerable attention from chemists. To date, several efficient synthetic strategies have been developed to access these compounds, including oxidation of 3-alkyl-2-oxindoles^[3] or 3-alkyl- indoles,^[4] electrophilic addition to 3-hydroxy-2-oxin- doles,^[5] and intramolecular cyclization of a-keto-amides.^[6] Despite their utility, these methods are often hampered by narrow substrate scope, harsh reaction conditions, or the need for specially designed precursors, thereby limiting their broad applicability. Additionally, following green transformations of isatins into complex molecules,^[7] the 3-alkylation of these commercially available and stable feedstocks has emerged as a viable route to 3-alkyl-3-hydr- oxy-2-oxindoles (Scheme 1a). In this regard, both the nucleophilic alkylation of isatins with organometallic reagents (Zn,^[8] In,^[9] Hg,^[10] Al,^[11] etc.^[12]) and the aldol reactions of isatins with aldehydes,^[13] ketones,^[14] and enol silyl ethers,^[15] as well as vinylogous aldol reactions,^[16] have been well established. In recent years, visible-light-induced radical-mediated alkylation of isatins via single-electron transfer (SET) or hydrogen atom transfer (HAT) pathways has emerged as a complementary approach for accessing these scaffolds.^[17] However, the introduction of a functionalized long-chain alkyl moiety remains scarce.

In recent years, N-radical cation-mediated C—C bond cleavage/functionalization of spiro pre-aromatic intermediates (spiro PAIs), such as dihydropyrazoles,^[18] dihydrotriazoles,^[19] and dihydroquinazolinones^[20] has gained attentions from chemists (Scheme 1b). This strategy enables the generation of alkyl radicals bearing a terminal heterocyclic aromatic ring with high atom economy, facilitating the formation of various chemical bonds. However, photocatalysts or stoichiometric amount of oxidants are usually required to generate the key N-radical cation intermediates. Given the inherent electronic properties of isatin (an electron-accepting arene)^[21] and dihydroquinazolinone (an electron-donating arene),^[22] we hypothesize that these two species can assemble to form an electron donor-acceptor (EDA) complex via π-π stacking. Subsequent photoexcitation of this EDA complex can facilitate the requisite single-electron transfer (SET) process, thereby obviating the need for external photocatalysts or stoichiometric oxidants. Herein, we report a visible-light-induced, copper-catalyzed alkylation of isatins with spiro dihydroquinazolinones via C—C bond cleavage, providing a new access to the 3-alkyl- 3-hydroxy-2-oxindoles (Scheme 1c). The key merits of this protocol include mild reaction conditions, operational simplicity, freedom from extra photocatalysts, and high atom economy.

Initially, N-Boc isatin 1a and spiro dihydroquinazolinone 2a were selected as model substrates to identify the optimal reaction conditions (Table 1). Gratifyingly, the reaction proceeded successfully in the presence of 5 mol% Cu(NO₃)₂ in dimethyl sulfoxide (DMSO) under 30 W Blue LEDs irradiation at room temperature for 6 h, yielding the desired product 3a in 77% isolated yield (Entry 1). Solvent screening revealed that polar aprotic solvent N,N-dimethyl- acetamide (DMAc) was also suitable, with DMSO still being the optimal solvent (Entry 2). Evaluation of catalysts demonstrated that other copper salts such as Cu(OTf)₂ and CuOTf were also effective, but afforded lower yields than Cu(NO₃)₂. Co(OAc)₂ gave a moderate yield of 3a, whereas Fe(OTf)₃ and Ga(OTf)₃ were less effective (Entries 3~5). A survey of light sources identified 465 nm blue LEDs as the optimal choice for this transformation. 390 nm purple of LEDs were found to be less effective, while 365 nm purple and 510 nm green LEDs proved ineffective (Entries 6 and 7). Finally, control experiments established that both light irradiation and an inert atmosphere (N₂) were indispensable for this transformation (Entries 8 and 9). Notably, in the absence of the copper catalyst, the ring-opening byproduct 5 was obtained in 57% yield as the sole product (Entry 10). We speculate that the copper catalyst activates the isatin carbonyl, enhancing the ketyl radical’s electrophilicity to facilitate the final radical-radical coupling.^[17d,17f]

With optimized conditions in hand, the generality and limitations of isatins were first evaluated using 2a as the alkylating reagent (Table 2). The N-protecting groups on the isatin scaffold significantly impacted this reaction. Substrates bearing various electron-withdrawing groups, such as acyloxy (Boc, Cbz), acyl (Ac, Bz) and sulfonyl (Ts) groups, were compatible under the standard conditions, affording the corresponding products 3a~3e in moderate to good yields. In contrast, isatins bearing an electron-donat- ing N-methyl group or unprotected N—H failed to deliver the desired products (3f and 3g). Attempts to promote these reactions using shorter wavelength irradiation at 365 and 390 nm or by adding 4CzIPN as an external photocatalyst were unsuccessful. This inactivity can be attributed to the diminished electron-accepting ability of these substrates, which disfavors the EDA complex formation. Subsequently, substituents on the aromatic ring of isatin were investigated. Both the position and electronic effects of substituents profoundly influenced the reaction efficiency. Isatins bearing electron-withdrawing or electron-donating substituents at the 4-position afforded the desired products 3h~3j in excellent yields. At the 5-position, substrates with electron-withdrawing or weakly donating substituents also furnished the desired products 3k~3m in moderate yields. However, the analogue with a strongly electron-donating group (OMe) failed to yield any desired product (3n). Substrates with electron-withdrawing or electron-donating substituents at the 6- or 7-position gave products 3o~3r in 46%~72% yields. 4-Bromo-5-methyl isatin reacted smoo-thly with 2a to deliver product 3s in 88% yield. While the 5,7-dimethyl isatin proved unreactive (3t), likely because the electron-donating groups attenuated the electron-accep- ting ability of the isatin core, thereby suppressing formation of a productive EDA complex. Remarkably, introduction of 4CzIPN (1 mol%) as an exogenous photocatalyst enabled the corresponding products 3n and 3t to be obtained in moderate yields, suggesting that an alternative photocatalytic cycle can overcome this limitation.

Next, the scope of spiro dihydroquinazolinones was examined with isatin 1a as the radical acceptor (Table 3). PAIs bearing electron-donating or electron-withdrawing groups on the aromatic ring all delivered the corresponding products in moderate yields (4a~4g), accompanied by approximately 10% the ring-opening byproduct 5 formed. Additionally, PAIs bearing different protecting group on the N-atom of amide, including N-Me, N-Ph and N-p-Tol, were also investigated. All of them successfully engaged in this C—C bond cleavage alkylation reaction, affording the desired products 4h~4j in moderate yields. Unfortunately, PAIs derived from cyclopentanone and 2-Me cyclopentanone (2k and 2l) both were invalid for this reaction, probably due to the less ring strain force or the rapid reverse cyclization. In these cases, the materials (2k and 2l) were almost recovered.

To demonstrate the practical utility of this protocol, a gram scale synthesis was performed (Scheme 2a). Impressively, when the model reaction was scaled up to 5.0 mmol (2a), the target product 3a was obtained in 65% yield (1.43 g) after extending the reaction time to 18 h. Further derivatizations of 3a were then explored (Scheme 2b). Under acidic conditions, deprotection of the N-Boc group in 3a proceeded smoothly, furnishing product 6 in 87% yield. To enhance the synthetic utility, a telescoped procedure integrating the condensation and ring-opening/coupling steps was designed (Scheme 2c). The telescoped protocol proceeded smoothly with cyclobutanone, affording the target product 3a in a moderate yield of 54%. To further expand the utility of this reaction, methyl benzoylformate 7 was used instead of isatin for this C—C bond cleavage/alkyla-tion reaction (Scheme 2d). When 1 mol% 4CzIPN was added as a photocatalyst, the anticipated product 8 was obtained in 45% yield.

To gain insight into the mechanism of this reaction, several control experiments were carried out (Figure 2). To probe for radical intermediates, radical-trapping experiments were conducted (Figure 2a). The addition of 1.0 equiv. of (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEM-PO) completely suppressed product formation, meanwhile the alkyl-TEMPO adduct 9 was isolated in 56% yield. Furthermore, addition of 1.0 equiv. of the radical inhibitor 2,6-di-tert-butyl-4-methylphenol (BHT) reduced the yield of 3a from 77% to 48%. These results collectively indicate that the transformation likely proceeds via a radical pathway and involves an alkyl radical intermediate generated from the ring opening of 2a. Next, light on/off experiments were carried out, indicating that continuous irradiation was essential for this transformation (Figure 2b). Stern-Volmer quenching experiments showed that 2a cannot quench the excited state of 1a, leading us to infer that the key pathway does not involve a diffusible excited state of free isatin (Figure 2c). This result therefore points to a ground-state EDA complex mechanism, wherein the adduct serves as the core unit for light absorption and charge transfer. Furthermore, UV-Vis absorption spectroscopy revealed that mixing 1a and 2a slightly enhanced absorption in the blue-light region (Figure 2d) and generated a distinct band at approximately 350 nm (see Supporting Information, Figure S4). This spectral feature, characterized by a weak absorption tail extending into blue wavelengths accounts for the complex’s visible-light-harvesting capability and photoactivity under blue LED irradiation. Job’s plot analysis confirmed a 1∶1 stoichiometry between 1a and 2a, consistent with a π-π-stacked EDA complex (Figure 2e). NMR titration experiments using the residual solvent peak of DMSO- d₆ (δ 2.50) as an internal standard revealed that upon the addition of 1a, the NMR signals of key aromatic proton (H-8) in 2a exhibited continuous and concentration-depen- dent chemical shift changes (Figure 2f). This regular displacement pattern confirms the existence of close, dynamic intermolecular interactions between the two species, providing further evidence for the formation of a ground- state EDA complex.

Moreover, density functional theory (DFT) calculations were conducted and provided further support for EDA complex formation (Figure 3). After full geometric optimization of the complex formed by substrates 1a and 2a, a stable EDA complex with a formation energy of －51.04 kJ/mol was obtained. Structural analysis revealed that 1a and 2a bind in a face-to-face manner (Figure 3a), indicating that the EDA complex possesses high stability and the driving force for its formation might be π-π interactions. Subsequently, frontier molecular orbitals (FMOs) analysis was performed on the EDA complex. The results demonstrated that the highest occupied molecular orbital (HOMO) is mainly localized on 2a (as donor), while the lowest unoccupied molecular orbital (LUMO) is predominantly distributed over 1a (as acceptor). This finding is consistent with the electron transfer process inferred earlier. In comparison with the unbound substrates, the HOMO-LUMO energy gap of the EDA complex is significantly reduced (Figure 3b). The theoretical UV-Vis absorption spectrum indicated that the absorption peak of the EDA complex undergoes a distinct red shift relative to that of the individual substrates, which is consistent with the experimental results (Figure 3c). Finally, the IRI analysis showed that the intermolecular interactions are primarily distributed in the planar region between 1a and 2a, with an interaction strength comparable to that of van der Waals interactions (Figure 3d). This observation further supports the possible existence of π-π interactions between the two molecules.

Based on relevant literature, experimental verification, and theoretical calculations, a probable EDA-mediated mechanism was proposed for this reaction (Scheme 3). Initially, isatin 1a and spiro dihydroquinazolinone 2a form an electron donor-acceptor (EDA) complex.^[21-22] Upon irradiation, this complex is excited and undergoes single- electron transfer (SET), generating the ketyl radical anion I and the N-radical cation II. The copper catalyst, acting as a Lewis acid, coordinates with the carbonyl oxygen of intermediate I to form intermediate III.^[17d,17f] This coordination enhances the electrophilicity of the carbonyl carbon, and stabilizes the intermediate, which is crucial for the subsequent radical cross-coupling step. The N-radical cation II then undergoes β-scission and deprotonation to produce the alkyl radical IV. Finally, radical cross-coupling between III and IV, followed by protonation, affords the final product 3a.

In conclusion, a sustainable EDA strategy for the C—C bond cleavage/coupling of spiro dihydroquinazolinones with isatins has been developed. This concise protocol requires no external photosensitizer or oxidant, and offers a mild, operationally simple, and atom-economical approach to synthesize 3-alkyl-3-hydroxy-2-oxindoles bearing a quinazolin-4(3H)-one motif in good yields with broad functional group tolerance. Gram-scale synthesis and derivatization further underscore the practical utility of this method. This work not only provides an efficient and green route to a privileged scaffold but also expands the toolkit for radical-mediated C—C bond cleavage and functionalization.