Organophosphorus compounds is recognized as one valuable functional scaffold in organic chemistry,^[1] which have excellent application prospects as materials,^[2] pharmaceuticals,^[3-4] and ligands.^[5] Among these compounds, phosphinoylindoles are basic elementary components of active natural products and materials in pharmaceuticals, demonstrating promising application prospects in functional molecules design (Figure 1).^[6] Therefore, the synthetic methodology of phosphinoylindoles has attracted considerable research attention in the scientific community.^[7] Despite these remarkable advancements, the development of novel and efficient synthetic methodologies for constructing phosphorylated indoles remains imperative.

Over the last few decades, numerous synthetic strategies have been developed for efficient synthesis of phosphonylated heterocycles. These approaches primarily encompass direct phosphonylation of heterocycles catalyzed by transition metals, visible light or electricity, and addition/ coupling reactions of heterocyclic compounds (Scheme 1, a).^[8] Furthermore, owing to their unique reactivity, isocyanides serve as highly efficient synthons for constructing heterocyclic frameworks.^[9] Hence, the cascade cyclization reactions involving isocyanides and phosphoryl groups have been widely applied to the synthesis of heterocyclic compounds containing phosphorus skeletons (Scheme 1b, path 1).^[10] In 2018, Yang’s group^[10g] have demonstrated a new, mild approach for the synthesis of 2-phosphinoyl- indoles through photoredox catalysis without the use of ex- ternal oxidant. In 2021, Yu’s group^[10j] developed a visible light-induced proton-coupled electron transfer strategy for the generation of phosphorus-centered radicals, via which a wide range of phosphorylated heterocycles were constructed. In 2022, Wu’s group^[10k] have successfully used single cobaloxime photocatalysis for the synthesis of phosphorylated heteroaromatics under mild redox-neutral conditions, and a variety of phosphine oxides directly added to isocyanides via P-radical addition and β-H elimination. In 2024, Studer’s group^[10m] presented a radical cascade addition cyclization sequence to access quinoline- based benzophosphole oxides from ortho-alkynylated aromatic phosphine oxides using various aryl isonitriles as radical acceptors. However, these reactions mainly proceeded through a radical mechanism, while the use of phosphorus central ions for the construction of phosphonylated heterocycles have been rarely reported.^[[8c-8f] In 2024, Xu’s group^[11] reported a silver catalyzed P-centered anion-initiated cascade reaction of functionalized arylisocyanides with H-phosphorus oxides (Scheme 1b, path 2). Although this approach has efficiently achieved the construction of phosphonated indoles, the reaction conditions remain suboptimal due to the reliance on noble metal catalysts and thermal activation.

Inspired by our research group’s interest in isocyanide chemistry and green synthesis,^[12] and aiming to expand the cascade cyclization reactions involving phosphoryl anions and isocyanides, we developed a metal-free annulation reaction that efficiently converts readily available 2-iso- cyanobenzaldehydes and aryl phosphine oxides into 2- phosphinoylindoles under mild conditions (Scheme 1c). This conversion pathway provides a new strategy for the synthesis of highly valuable phosphorylated heterocycle compounds.

Initially, the cyclization reaction of 2-isocyanobenzalde- hyde 1a and diphenyl phosphine oxide 2a was chosen as a model to screen the reaction conditions. We firstly examined the solvents and it was found that MeCN was used to achieve high yield (Table 1, Entries 1~4). When MeCN was used as the solvent, a series of bases including 1,4- diazabicyclo[2.2.2]octane (DABCO), t-BuOK, tetramethyl- guanidine (TMG) and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) were tested (Table 1, Entries 4~8), and the results showed that the desired product 3aa was obtained in 73% yield in the presence of DBU (Table 1, Entry 4). To further optimize the reaction, the temperature was screened, and the yield of 3aa was found to decrease at both elevated and reduced temperature. Furthermore, after determining the optimal base loading to be 2.5 equiv., the influence of substrate equivalents on the reaction efficiency was subsequently investigated. Notably, when substrate 2a was employed at 2.5 equiv., the reaction demonstrated enhanced productivity, achieving a product yield of up to 85% (Table 1, Entries 10~13). Finally, a series of control experiments were conducted to investigate the effects of oxygen and water on the reaction. The results indicated that water exhibited no significant effect on the reaction process, whereas the presence of oxygen decreased the product yield. Therefore, the reaction conditions in Entry 13 was proved to be the optimal.

With the optimized reaction conditions, we firstly examined the substrate scope of the H-phosphorus oxides (Table 2). When para-substituted aryl phosphine oxides bearing electron-withdrawing groups were employed, the corresponding 2-phosphinoyl indoles were obtained in 83%~88% yields (3ab~3ad). In comparison, substrates with electron-donating groups (OMe, Me) afforded products 3ae~3af in slightly lower yields. In addition, the scope could be extended to 3,5-dimethyl-substituted aryl phosphine oxide (3ag). The product 3ah containing 2-naphthyl was synthesized in moderate yield. However, alkyl phosphine oxides failed to generate the desired products under the same conditions, indicating the necessity of aromatic substrates for this transformation. Furthermore, the target product 3ai was also obtained in 81% yield when the R² and R³ groups of the H-phosphorus oxide 2i were distinct. Subsequently, the substrate scope of 2-isocyano- benzaldehydes 1 were investigated (Table 2). Substrates with both electron-withdrawing and electron-donating substituents on the aromatic ring afford the corresponding products in excellent yields (3ba~3fa). Additionally, when the position and number of substituents on the aryl ring of the 2-isocyanobenzaldehydes were systematically varied, the corresponding products were obtained in good yields (Table 2, 3ga~3ia). Finally, when the para-position of the aryl ring in substrate 1a was substituted with a 1-naphthyl or 2-thienyl group, the corresponding products were obtained in 76%~83% yields (Table 2, 3ja~3ka). Notably, product 3la, containing a 1,3-benzodioxole motif, was obtained in 88% yield.

To evaluate the synthetic applicability of this cascade reaction, gram-scale syntheses of 3aa was performed. Under the standard conditions, 1.04 g of product 3aa was obtained in 82% yield (Scheme 2a). Subsequently, several synthetic transformations were carried out. At first, The thiolation of P=O bond of 3aa was conveniently realized by using Lawessonʼs reagent, affording corresponding pro- duct 4 yield of 75%.^[13] In the presence of potassium hydroxide, compound 3aa underwent nucleophilic substitution with allyl bromide in N,N-dimethylformamide (DMF), affording the allylation product 5 in 91% yield.^[14] Finally, the reduction of phosphinoylindole 3aa using HSiCl₃ afforded 6 in 82% yield (Scheme 2b).^[7d]

Subsequently, a series of control experiments were carried out to gain insights into the reaction mechanism. No desired product 3aa was observed when the cyclization reaction was treated in the absence of base. When 2,2,6,6- tetramethylpiperidine-1-oxyl (TEMPO, 5.0 equiv.) and 2,6-di-tert-butyl-4-methylphenol (BHT, 5.0 equiv.) were added as radical inhibitors under standard conditions, they could not markedly influence the yield of 3aa (Table 3). These results indicate that a P-centered radical pathway could most likely be ruled out in this transformation. When H-phosphorus oxide 2j containing a large site-blocking group was used as a substrate, bisphosphonylaminomethane 7 was obtained in 70% yield, while the corresponding 2-phosphonyl indole product could not be detected (Scheme 3). These results indicate that under steric hindrance, iso-cyanide 1a undergoes anionic 1,1-bisphosphination^[16] competing with cyclization, further confirming the absence of a radical pathway in this synthetic strategy.

Based on the above results and previous reports in the literature,^[15b,16] a plausible reaction mechanism is proposed (Scheme 4). Initially, the presence of DBU facilitates the deprotonation process of compound 2a, producing a P- center anion. Then, the nucleophilic addition of isocyanobenzaldehyde 1a leads to the imine intermediate A. Subsequently, intermediate A would be attacked by another H-phosphorus oxide 2a to form intermediate B, followed by a [1,2]-Phospha-Brook rearrangement to generate intermediate C. Intermediate C then undergoes cyclization and protonation to afford intermediate D. Finally, elimination of a molecule of diphenylphosphinic acid from intermediate D yields the final product 3aa. The molecular weight of the intermediates A, D and diphenylphosphinic acid were detected by HRMS.

In conclusion, we have established a practical and scala-ble synthetic strategy for constructing 2-phosphinoylindoles through base-promoted cascade reactions of isocyanobenzaldehydes with phosphorus oxides. This protocol features mild conditions, excellent functional group tolerance, and scalability. This approach represents a green and sustainable alternative to metal-catalyzed, photochemical, or electrochemical phosphorylation strategies. Extensive experimental validation confirms the versatility of this method, which offers significant advantages in phosphine ligand synthesis, biorthogonal chemistry, and drug discovery.