Spirocyclic compounds are of great significance in medicinal chemistry due to their unique three-dimensional structures and distinctive biological and pharmaceutical activities.^[1] Among them, spiroindolines are inherently high-value motifs of significant biological importance, found in numerous natural products and biologically active substances, such as kopsinine,^[2] Sky kinase inhibitors,^[3] and spiroindimicin^[4] (Figure 1).

In the past decades, the synthesis of spiroindolines has attracted extensive attention. One of the most straightforward strategies is the dearomatization of indoles. In 2010, the You group^[5] realized an Ir-catalyzed intramolecular C-3 allylic alkylation of indoles, affording highly enantioenriched spiroindolenine derivatives. In a subsequent reduction step, the spiroindolenine could be reduced to the corresponding spiroindoline. After this seminal work, various synthetic strategies employing Ir,^[6] Rh,^[7] Au,^[8] Cu,^[9] Pd,^[10] and Ag^[11] have been successively reported for the construction of spiroindolines (Scheme 1a).^[12] However, most of these methods still require an additional reduction step, which compromises step economy and limits their practical utility. Palladium-catalyzed Heck reactions play an important role in the construction of heterocyclic com- pounds.^[13] The intramolecular Heck reaction of an aryl halide with a tethered terminal alkene is one of the alternative methods for the synthesis of spiroindolines.^[14] The reaction proceeds through a spiropalladacycle intermediate A, which is subsequently trapped by nucleophiles such as benzynes,^[15] α-diazocarbonyl compounds,^[16] CH₂Br₂,^[17] arynes,^[18] and diaziridinone^[19] to construct the spiroindoline framework (Scheme 1b). In contrast to these reported Heck spirocyclizations that rely on external nucleophiles to trap the spiropalladacycle intermediate, our method utilizes the intrinsic N—H moiety as the internal nucleophile, bypassing the need for an additional trapping reagent. However, to the best of our knowledge, the synthesis of spiroindolines via the intramolecular Heck reaction of alkenes has not been realized. Herein, we report a palladium-catalyzed intramolecular Heck/amination of unactivated alkenes that proceeds through an N,P-containing spiropalladacycle intermediate B for the synthesis of spiroindolines. The reaction employs the intramolecular N—H moiety as the nucleophile without the need for external nucleophiles, proceeding via direct intramolecular N—H insertion to afford spiroindolines. This method affords a series of spiroindolines in good to moderate yields, and preliminary asymmetric attempts have also been conducted (Scheme 1c).

We initiated our studies on the palladium-catalyzed Heck/amination reaction using ortho-iodophenol-derived isopropenyl diphenylamine ether 1a as a model substrate (Table 1). Initially, various ligands were screened in the presence of Pd(dba)₂ as the palladium source, Cs₂CO₃ as the base, and toluene as the solvent at 110 ℃ for 13 h. Under these conditions, PPh₃ delivered the desired product in 48% yield (Table 1, Entry 1). Among the ligands examined, the bulky ligands S-Phos and DavePhos afforded 47% and 34% yields, respectively (Table 1, Entries 2, 3). PCy₃ gave a yield of 16% (Table 1, Entry 4). The electron-rich (4-MeO- C₆H₄)₃P and (4-MeC₆H₄)₃P gave moderate yields (Table 1, Entries 5, 6). The electron-withdrawing ligand (4-CF₃- C₆H₄)₃P proved superior, affording the product in 71% yield (Table 1, Entry 8). Subsequently, the palladium source was examined using (4-CF₃C₆H₄)₃P as the optimal ligand. When various palladium catalysts were evaluated, Pd(dba)₂ remained the most effective, whereas Pd(PPh₃)₄, PdCl₂, and Pd(PPh₃)₂Cl₂ led to significantly lower yields (Table 1, Entries 9~11). Further optimization of the base showed that Cs₂CO₃ and K₂CO₃ gave comparable results (Table 1, Entry 12), while CsOAc resulted in a lower yield (Table 1, Entry 13). When the organic base Et₃N was employed, no desired product was observed (Table 1, Entry 14). Finally, lowering the reaction temperature to 90 ℃ resulted in a decreased yield of 55% (Table 1, Entry 15), confirming that 110 ℃ was optimal. Control experiments were also conducted (Table 1, Entries 16, 17). Omitting either the palladium source (Pd(dba)₂) or the phosphine ligand (4-CF₃- C₆H₄)₃P resulted in no reaction, indicating that both components are essential for the transformation.

With the optimized reaction conditions in hand (Table 1, Entry 8), the substrate scope suitable for this strategy was explored (Scheme 2). The reactions proceeded smoothly when the N-phenyl substituent was changed to N-naphthyl, albeit in diminished yield (53%), giving product 2b. When the N-substituent was changed to alkyl groups, such as N-benzyl and N-methyl, the reactions proceeded smoothly, affording the corresponding products 2c and 2d in moderate yields. When changing the substituent on Ar¹, both the elec- tron-donating methyl group and the electron-withdrawing chloro group allowed the reaction to proceed smoothly, affording the desired products 2e~2g. However, when the electron-withdrawing chloro group was located at the meta-position of the N-substituent, the yield of 2g was lower (41%) than that at the para-position (2f, 71%). Subsequently, the applicability of various substituents on Ar² was examined to the reaction. When the methoxy group was located at the ortho-position relative to iodine (1h), although the steric hindrance increases, the reaction still affords the corresponding target product 2h in 67% yield. When a methyl group was added at the meta-position relative to iodine (1i), the desired product 2i was obtained in 77% yield. When a bulky tert-butyl group or electron-with- drawing chloro and trifluoromethyl groups were introduced at the meta-position, the yields of products 2j~2l were all moderately decreased compared to the model substrate. Introducing electron-donating methoxy (1m) and electron-withdrawing chloro (1n) substituents at the para-posi- tion relative to iodine gave products 2m and 2n in 82% and 64% yields, respectively. However, changing Ar² to pyridine (1o) resulted in poor conversion, primarily due to the electron-withdrawing nature of the pyridine ring and possible palladium coordination by the pyridyl nitrogen. Under the reaction conditions, replacing iodine with bromine reduced the yield of 2a to 11%, presumably because aryl bromides undergo slower oxidative addition than aryl iodides.

In view of the high potential of chiral spiroindolines in drug discovery and as chiral ligands, the replacement of (4-CF₃C₆H₄)₃P with chiral phosphine ligands was explored to access these spirocyclic frameworks. After screening a series of chiral phosphine ligands and solvents, the optimal result to date was obtained using the N-P ligand L*, which gave the target product 2a in 87% yield with 65% ee (Eq. 1).

On the basis of experimental observations and known literature, a reasonable catalytic cycle is proposed as illustrated in Scheme 3.^[13a] Initially, oxidative addition of 1 to the Pd(0) complex generates arylpalladium species I. Subsequently, I undergoes intramolecular 5-exo-cyclization to afford alkylpalladium intermediate II. Then, deprotonation of the N—H bond by the base generates a nitrogen anion, which subsequently attacks the palladium center to form spiropalladacycle III. Finally, reductive elimination from III produces the corresponding product 2 and regenerates the Pd(0) catalyst for the next catalytic cycle.

In summary, we have developed a palladium-catalyzed intramolecular Heck/amination for the synthesis of spiroindolines using the intrinsic N—H moiety as the nucleophile. This method offers a step-economical approach to spiroindoline frameworks without the need for external nucleophiles. Further studies on the asymmetric synthesis of spiroindolines are currently underway in our laboratory.