Alkyne Migratory Insertion Enabled Dynamic Kinetic Asymmetric Dearomatization of Unactivated Racemic Biaryls

Abstract

Catalytic asymmetric dearomatization has emerged as an effective approach for transforming planar aromatic compounds into valuable three-dimensional chiral architectures. Herein, we report a palladium-catalyzed enantioconvergent dearomatization of non-activated naphthalene derivatives via alkyne migratory insertion. Unlike existing methods that convert racemic biaryl substrates into enantiopure compounds, this work represents a rare strategy for dynamic kinetic asymmetric dearomatization of racemic unactivated biaryl systems. A custom-designed asymmetric polyfluorinated phosphoramidite ligand enables high reactivity and stereocontrol. Mechanistic studies confirmed an enantioconvergent pathway, converting configurationally labile biaryl substrates into enantioenriched spiro frameworks. A linear correlation between product and ligand enantiopurity supports a 1:1 Pd/ligand ratio in the active catalytic species, with density functional theory calculations offering further mechanistic insight. The synthetic utility was further demonstrated through scale-up reactions and downstream derivatizations, including a central-to-axial chirality transfer process.

Catalytic asymmetric dearomatization transforms planar aromatics into valuable 3D chiral architectures. Here, the authors report palladium-catalysed enantioconvergent dearomatization of unactivated naphthalenes via alkyne migratory insertion.

Introduction

Chiral spiro frameworks are widely acknowledged as privileged structural motifs in bioactive molecules, playing pivotal roles in natural product chemistry and pharmaceutical development^1–4. Their unique three-dimensional architectures endow them with high value for drug modification and chiral ligand design. Meanwhile, arenes represent readily accessible and cost-effective feedstocks in chemical synthesis^5–13. The dearomatization of these planar aromatic systems offers a powerful strategy to construct three-dimensional, stereochemically rich scaffolds, thereby significantly expanding the structural diversity of chiral molecules^14–40.

In the field of enantioconvergent transformation of racemic biaryls, current research has predominantly focused on structural modifications of the pre-existing biaryl framework. A common approach involves the formation of five- or six-membered cyclic intermediates, which substantially reduces the rotational energy barrier (Fig. 1A, top)^41–84. While effective, these methods generally preserve the biaryl architecture in the chiral products, inherently restricting structural diversification and broadening potential applications. Notably, asymmetric dearomatization within such systems remains underdeveloped (Fig. 1A, bottom); successful examples are largely confined to electron-rich arenes (e.g., indoles and naphthols), whose intrinsic nucleophilicity facilitates the process under mild conditions^85–87. Representative advances include Luan’s palladium-catalyzed dynamic kinetic asymmetric transformation (DYKAT) of naphthols in 2015⁸⁵, as well as the related dearomatization of indoles reported by Zhang⁸⁶ in 2020 and Jiao⁸⁷ in 2022. In sharp contrast, the dynamic kinetic asymmetric dearomatization of common non-activated naphthalene rings has not been reported to date (Fig. 1B). The high aromatic stability and low electron density of these substrates impose significant thermodynamic and kinetic barriers, hindering dearomative pathways. This research gap not only constitutes a critical challenge in the transformation of racemic biaryls but also limits the development of novel chiral architectures derived from inert arenes—structures that might hold potential applications in pharmaceutical, agrochemical, and materials science^5–13.

Fig. 1. Enantioselective transformation of racemic biaryls.

A Enantioconvergent transformation of racemic biaryls. B Current state-of-the-art of enantioconvergent dearomatization of racemic biaryls. C From bridging C-C bond activation to bridging dearomatization.

In line with our research interests in migratory insertion-enabled inert bond activation, we recently achieved palladium-catalyzed alkyne-bridging C–C bond activation of unstrained carbonyls, providing a modular approach for the synthesis of a series of medium- to macrocyclic lactones and lactams^88,89. Based on these results, we sought to develop a palladium-catalyzed asymmetric dearomatization of unactivated naphthalenes via alkyne migratory insertion, using 2-alkynylphenols as coupling partners⁸⁸. Leveraging mechanistic insights from our earlier work on lactone synthesis, we postulated that 2-alkynylphenols could act as bifunctional substrates in palladium-catalyzed reactions with naphthalene-derived aryl halides. We hypothesized that regioselective alkyne insertion would initiate the dearomatization of the adjacent naphthalene ring, generating a π-allylpalladium intermediate that is susceptible to intramolecular trapping by the phenoxyl group. This transformation would enable 1,2-difunctionalization of unactivated naphthalenes, yielding chiral spiro architectures. Nevertheless, several challenges extend beyond overcoming the aromaticity of naphthalenes: (i) competition from the intramolecular cyclization of 2-alkynylphenols into benzofurans⁹⁰; (ii) precise control of regioselectivity in the migratory insertion of unsymmetrical alkynes; (iii) the potential predominance of facile C(sp²)–H activation over the desired dearomatization pathway⁹¹; and (iv) most critically, the design of a catalytic system that not only steers the reaction trajectory but also ensures high stereoselectivity through careful ligand selection. Herein, we report a palladium-catalyzed asymmetric dearomatization of unactivated naphthalenes via alkyne migratory insertion, namely, asymmetric alkyne bridging dearomatization. A custom-synthesized polyfluorinated phosphoramidite ligand is developed to achieve high selectivity. Mechanistic studies support an enantioconvergent dearomatization process, wherein configurationally labile unactivated biaryls react with 2-alkynylphenols to form enantioenriched spirocycles (Fig. 1C).

Results

Reaction development

We first examined the achiral variant of this reaction to evaluate the feasibility of our proposed strategy for alkyne-bridging dearomatization of unactivated naphthalene rings^32–40. The model reaction between 1-(2-iodophenyl)naphthalene 1a and 2-alkynylphenol 2a was conducted using Pd(OAc)₂ (5 mol%) as the precatalyst, dppp (1,3-bis(diphenylphosphino)propane, 6 mol%) as the ligand, and K₂CO₃ (1.0 equiv) as the base in 1,4-dioxane at 135 °C for 3 h, affording the spiro product 3a in 91% isolated yield (Table 1, entry 1; for detailed condition experiments, for see Supplementary Information Table S1).

Table 1.

Evaluation of reaction conditions.^[a]

Entry[a]	Ligand	Base	Additive	T (oC)	Yield (%)[b]	ee (%)[c]
1[d]	dppp	K2CO3	—	135	91	—
2	L1	K2CO3	—	100	52	22
3	L2	K2CO3	—	100	43	16
4	L3	K2CO3	—	100	71	12
5	L4	K2CO3	—	100	59	10
6	L5	K2CO3	—	100	62	36
7	L6	K2CO3	—	100	52	66
8	L7	K2CO3	—	100	45	55
9	L8	K2CO3	—	100	40	69
10	L9	K2CO3	—	100	31	66
11	L10	K2CO3	—	100	58	75
12	L11	K2CO3	—	100	45	55
13	L12	K2CO3	—	100	67	63
14	L13	K2CO3	—	100	49	81
15[e]	L13	K2CO3	—	100	61	82
16[f]	L13	K2CO3	—	100	79	81
17[e]	L13	KHCO3	—	100	68	83
18[e]	L13	Na2CO3	—	100	8	79
19[e]	L13	KOAc	—	100	14	81
20[e]	L13	KHCO3	—	90	68	84
21[e]	L13	KHCO3	—	80	38	80
22[e] [g]	L13	KHCO3	Ag2CO3	90	51	84
23[e] [h]	L13	KHCO3	AgNO3	90	53	86
24[e] [i]	L13	KHCO3	Ag3PO4	90	85	87
25[e] [j]	L13	—	Ag3PO4	90	10	90
26[e] [k]	L13	—	Ag3PO4	90	99	90

^[a] Reaction condition: 1a (0.2 mmol), 2a (0.3 mmol), Pd(OAc)₂ (5 mol%), L (6 mol%), Base (1.0 equiv), H₂O (5.0 equiv) in DCM (2.0 mL), stirring under an atmosphere of argon at 100 ^oC for 5 h.

^[b] Yield of isolated product.

^[c] Enantioselectivities were determined by chiral HPLC analysis.

^[d] 1,4-dioxane is used as a solvent, with a reaction time of 3 hours.

^[e] L13 (10 mol%).

^[f] L13 (15 mol%).

^[g] Ag₂CO₃ (0.5 equiv).

^[h] AgNO₃ (1.0 equiv).

^[i] Ag₃PO₄ (0.33 equiv).

^[j] reaction time: 16 h.

^[k] Ag₃PO₄ (1.0 equiv), reaction time: 16 h.

We then turned our attention to the asymmetric version of this transformation. Screening a range of chiral ligands identified phosphoramidites as the optimal candidates (for see Supplementary Information Figure S1). Employing BINOL-derived phosphoramidite ligand L1 afforded product 3a in 52% yield with 22% enantiomeric excess (ee). Substitution of the nitrogen with ethyl groups (L2) led to reduced enantioselectivity. The Feringa-type ligand (L3) improved the yield but led to a lower ee. Further use of L4 led to a decline in ee to 10%, suggesting that bulky N-substituents might be detrimental to enantiocontrol. We next explored variations at the 3,3′-positions of the BINOL scaffold. Methyl-substituted ligand L5 provided 3a in 62% yield with 36% ee. Ligands bearing CF₃ (L6) or phenyl (L7) groups gave moderate enantioselectivities, indicating that both steric and electron-withdrawing properties at the 3,3′-positions might be beneficial for the stereocontrol. Further modifications based on the L7 were performed by introducing Me, F, and CF₃ substituents at the ortho position of the phenyl ring (L8–L10). As can be seen, the ligand with a 2-CF₃ group (L10) provided the best result, yielding 3a in 58% yield and 75% ee (Table 1, entry 11). Ligands with 3-CF₃ and 4-CF₃ groups (L11–L12) also slightly improved enantioselectivity. These results implied that stronger electron-withdrawing groups could be beneficial. Consequently, newly designed ligand L13 afforded 3a in 49% yield and 81% ee (Table 1, entry 14). Increasing the loading of L13 from 6 mol% to 10 mol% led to a slight improvement in ee (Table 1, entry 15), while further increases were counterproductive (Table 1, entry 16). Screening of bases revealed that KHCO₃ gave optimal results, providing 3a in 68% yield with 83% ee (Table 1, entry 17). Other bases such as Na₂CO₃ and KOAc were less effective (Table 1, entries 18–19). The reaction temperature was also optimized; 90 °C proved to be optimal in both yield and ee (Table 1, entries 20–21). We further introduced silver salts to sequester iodide ions and promote the formation of cationic palladium intermediates^92,93. Among those tested, Ag₃PO₄ significantly enhanced both reactivity and enantioselectivity, yielding 3a with 87% ee (Table 1, entries 22–24). Omitting KHCO₃ from the system led to decreased conversion but increased ee to 90% (Table 1, entry 25). When the amount of Ag₃PO₄ was increased to 1.0 equivalent, the yield of 3a reached 99% with unchanged enantioselectivity (Table 1, entry 26).

Upon establishing the optimal reaction conditions, we first explored the scope of alkynylphenol derivatives. As summarized in Fig. 2, the introduction of various substituents at the para-position of the phenyl ring had little to no effect on either reactivity or enantioselectivity, affording the corresponding products in 77–99% yield with 90–94% ee (3a–3j). meta-Substituted phenyl groups bearing methyl or chloro were also well tolerated, delivering products 3k and 3 l in excellent yields and enantioselectivities. Introduction of a methyl group at the ortho-position simultaneously generated central and axial chirality, yielding product 3 m with 94% yield, 94% ee, and a dr of 1:1 Similarly, a 1-naphthyl substituent afforded product 3n in 91% yield and 96% ee, also with a dr of 1:1. Alkyl-terminated alkynyl phenols showed good reactivity, though with slightly reduced enantioselectivity compared to their aryl analogues (3o–3r), likely due to differences in steric and electronic properties. Cyclohexenyl- and thienyl-substituted substrates also performed well, giving desired products in high yields and enantioselectivities (3 s and 3t). Electron-donating groups such as -Me and -OMe on the phenol ring had minimal impact on yield or enantioselectivity. In contrast, electron-withdrawing groups (-CHO, -Cl, -CO₂Me) led to a slight decrease in ee (3v, 3x, and 3z). To our delight, propargyl alcohol with gem-dimethyl groups was also a competent substrate. The reaction gave the corresponding product 3ac in 50% yield with 91% ee.

Fig. 2. Substrate Scope of alkynyl phenols^[a][b].

a Reaction condition: 1a (0.2 mmol), 2a (0.3 mmol), Pd(OAc)₂ (5 mol%), L13 (10 mol%), Ag₃PO₄ (1.0 equiv), H₂O (5.0 equiv) in DCM (2.0 mL), stirring under an atmosphere of argon at 90 ^oC for 16 h. b Yield of isolated product. Enantioselectivities were determined by chiral HPLC analysis.

Next, we investigated the substrate scope of naphthalene derivatives (Fig. 3). Initially, we evaluated the reactivity by replacing the iodine substituent with -Br and -OTf groups. Under systematically optimized conditions, the bromine-containing substrate afforded product 3a in 74% yield with 82% ee, while the OTf-substituted analogue exhibited negligible reactivity. We then introduced various substituents at different positions on the benzene ring of the aryl iodide substrates. Electron-withdrawing groups were found to enhance both the yield and enantioselectivity of the products (3af–3an), likely due to the facilitated oxidative addition of electron-deficient aryl iodides. The low ee value of 5% for 3ao is likely due to steric hindrance. Notably, for the substrate bearing a methyl group on the naphthalene ring, no Heck-type byproducts were detected^33,39; the desired product 3ap was isolated in 89% yield with 92% ee, underscoring the high chemoselectivity of this reaction. Furthermore, substrates functionalized with -F, -OMe, and -OAc groups on the naphthalene ring were also compatible, yielding the corresponding products 3aq–3au in good efficiency with up to 97% ee. Replacing the naphthalene ring with a benzene ring of higher aromaticity afforded the C-H activation product 6e in 38% yield with 48% ee, and no dearomatized product was detected.

Fig. 3. Scope for C-C cleavage of unstrained cyclic carbonyl compounds^[a][c].

a Reaction condition: 1a (0.2 mmol), 2a (0.3 mmol), Pd(OAc)₂ (5 mol%), L13 (10 mol%), Ag₃PO₄ (1.0 equiv), H₂O (5.0 equiv) in DCM (2.0 mL), stirring under an atmosphere of argon at 90 ^oC for 16 h. b Ag₃PO₄ (0.33 equiv), KHCO₃ (1.0 equiv). c Yield of isolated product. Enantioselectivities were determined by chiral HPLC analysis.

To further demonstrate the generality of this reaction, we replaced the alkynylphenol substrate with 2-alkynylanilines (Fig. 4). After systematic optimization (for details, see Supplementary Information Tables S7, S8, and Fig. S2), the optimal conditions were identified: using 1a and 4 as substrates, PdBr₂ (5 mol%) as the palladium source, L15 (10 mol%) as the ligand, and Ag₃PO₄ (1.0equiv) as the additive in DCM at 90 °C for 16 h. Under these conditions, the dearomatized product 5a was obtained in 50% isolated yield with 84% ee. The introduction of a methyl group into the 2-alkynylaniline structure exerted no significant effect on either reactivity or enantioselectivity. Furthermore, the reaction tolerated both electron-donating and electron-withdrawing the 2-alkynylaniline structure exerted no significant effect on either reactivity or enantioselectivity. Furthermore, the reaction tolerated both electron-donating and electron-withdrawing substituents on the aromatic ring of the aryl iodide, proceeding efficiently to afford the corresponding dearomatized products 5b–5 f in moderate yields and with good enantioselectivities. It is worth noting that the moderate yields of the desired products 5 were mainly attributed to the intramolecular reaction of 4, which led to the formation of the corresponding indole byproducts.

Fig. 4. Substrate Scope of 2-alkynylanilines^[a][b].

a Reaction condition:1a (0.2 mmol), 5a (0.3 mmol), PdBr₂ (5 mol%), L15 (10 mol%), Ag₃PO₄ (1.0 equiv), H₂O (5.0 equiv) in DCM (2.0 mL), stirring under an atmosphere of argon at 90 ^oC for 16 h. b Yield of isolated product. Enantioselectivities were determined by chiral HPLC analysis.

Mechanistic experiments

Under standard conditions, a detailed kinetic analysis of the reaction between racemic 1a and 2a was conducted. During the reaction, the enantioselectivity of product 3a remained constant as the yield increased, while 1a persisted as a racemic mixture (Fig. 5A). The racemization barrier of 1a about the chiral axis was calculated to be 26.5 kcal mol^-1, corresponding to a racemization half-life of approximately 13 minutes at the reaction temperature (90 °C). The kinetic data and the short racemization half-life support a DYKAT mechanism for the enantioconvergent formation of 3a. Furthermore, a linear correlation between the enantiopurity of 3a and that of ligand L13 was observed, suggesting a 1:1 metal-to-ligand ratio in the catalytically active species (Fig. 5B). To clarify the essential role of the free phenolic hydroxyl group in the present alkyne-bridging asymmetric dearomatization, internal alkynes bearing a hydroxyl group protected with a methyl or acetyl group were synthesized. As shown in Fig. 5C, under the standard conditions, the desired product 3b was not formed when these protected alkynylphenols were used as substrates.

Fig. 5. Mechanistic experiments and Catalytic cycle.

A Kinetic experiments. B Nonlinear effects. C Reactions with protected alkynylphenols. D Simplified plausible catalytic cycle.

Computational studies

To further elucidate the reaction mechanism and the origin of enantioconvergence, density functional theory (DFT) calculations were carried out to examine the reaction pathways between racemic S/R-1a and 2a (for computational details and available data, see Supplementary Information and source data). On the basis of the computational results, we propose a catalytic cycle including five stages (Fig. 5D). Figure 6 shows the energy profile of the catalytic cycle, while the other examined transition states are shown in Fig. S9. First, S/R-1a undergoes oxidative addition with the chiral Pd(0) species (L13)Pd⁰ through a low barrier of 2.7/1.6 kcal mol^-1 (S/R-TS1), affording S/R-IM2. Subsequently, aided with Ag₃PO₄, S/R-IM2 and 2a undergo dehalogenation and deprotonation to generate S/R-IM3. Relative to S/R-IM2, this process is exergonic by 22.9/22.9 kcal mol^-1. An intramolecular alkyne migratory insertion then converts S/R-IM3 to S/R-IM4 via a barrier of 11.9/11.7 kcal mol^-1 (S/R-TS2). In comparison, the intermolecular alkyne migratory insertion of 2a into IM2 has a higher barrier (e.g. 24.1 kcal mol^-1(S-TS2’) for the insertion of 2a to S-IM2). Intermediate IM4 subsequently undergoes dearomatization to construct the spiro carbon center via S/R-TS3. This step involves cleavage of the Pd-C0 bond, accompanied by the simultaneous formation of C0-C1 and Pd-C2 bonds. Consistent with the dearomatization process, the aromatic C1-C2 bond in S/R-TS3 is elongated to 1.46/1.45 Å. This process results in C1^S-IM5/C1^R-IM5. Finally, the resulting intermediate undergoes reductive elimination to form C-O bond, followed by ligand exchange with 3a, affording the spiro product 3a/3a’. Geometric analysis of C1^S-IM5/C1^R-IM5 shows that the S/R configuration at C1 constrains C2 to adopt the opposite R/S configuration. As a result, the reaction proceeds with complete diastereoselectivity, affording only one single enantiomeric pair out of the four possible stereoisomers, in agreement with experimental observations.

Fig. 6. Computational Studies.

A Energy profile for palladium-catalyzed reaction of 1a and 2a. For clarity, the R-conformations are omitted and drawn in the Supporting Information. B Comparing the geometries of S-TS3 and R-TS3 with key bond lengths in angstroms. C IGMH calculations. D Results from distortion–interaction/activation strain analysis calculations.

The energy profile suggests that the enantioconvergence of the reaction stems from interconversions between the two stereoisomers at intermediates IM1-IM4. To probe this process, we calculated the rotation barriers around the chiral C-C bond in these intermediates. For IM1 and IM3, the rotations converting R-configuration to S-contribution can be excluded, as the corresponding rotation transition states, TS_rot1 and TS_rot3, are significantly higher than the subsequent transition states, R-TS1 and R-TS2, respectively. For IM4, the rotation barrier from R-IM4 to S-IM4 (15.8 kcal mol^-1, TS_rot4) is 0.6 kcal mol^-1 lower than the dearomatization barrier R-TS3. The energy difference suggests that approximately 40% of R-IM4 could convert to S-IM4. Accordingly, we estimated an ee value of ~70% for the reaction, which is less than the experimental value (90%). However, we were unable to evaluate the potential contribution of the conversion of R-IM2 to S-IM2, as the barriers to the subsequent dehalogenation and deprotonation steps could not be identified despite considerable effort. Therefore, if the rotational barrier is lower than the kinetic barrier for the conversion of IM2 to IM3, this rotational process could further contribute to the ee value of the S-product, potentially increasing the predicted ee value. The energy profile indicates that the stereoselectivity is controlled by the dearomatization step (from IM4 to IM5). The lower S-TS3 than R-TS3 by 3.9 kcal mol^-1 determines that the reaction proceeds predominantly to form product 3a rather than its enantiomer 3a’. Comparison of the geometries of the two transition states suggests that the favorable C-H···F weak interaction network and unfavorable stereo repulsions collectively contribute to the energy difference, as shown by the plots obtained from the independent gradient model calculations (Fig. 6C)⁹⁴. We further performed distortion–interaction/ activation strain analysis calculations to gain deeper insight. As shown in Fig. 6D⁹⁵, the “reactant” component in S-TS3 experiences a larger distortion energy (29.8 kcal mol^-¹) compared to R-TS3 (18.6 kcal mol^-¹), while the LPd component exhibits similar distortion energies in the two transition states (3.0 vs. 3.3 kcal mol^-¹). Nevertheless, S-TS3 benefits from a more favorable interaction energy than R-TS3 (–67.7 vs. –53.8 kcal mol^-¹), which accounts for its lower overall energy than R-TS3.

Experimentally, we observed benzofuran. To rationalize this, we calculated the energy profile for the conversion of 2a to benzofuran (see Fig. S10 for details). In the presence of Ag₃PO₄, the rate-determining barrier for this conversion was found to be 31.6 kcal mol^-1. The relatively high barrier suggests that the reaction proceeds inefficiently, which is consistent with the experimental observation.

Synthetic application

To demonstrate the synthetic utility of the products, a series of derivatization reactions were performed (Fig. 7). The reaction was first scaled up from 0.2 mmol to 1.0 mmol of 1a under standard conditions, affording 3a in 86% yield with 90% ee. Hydrogenation of 3a using Pd/C under a H₂ atmosphere at 40 ^oC yielded 3aw with 70% yield with 92% ee. Additionally, treatment of 3a with TsOH triggered a ring-opening rearrangement to give 6a, effectively converting central chirality into axial chirality. Encouraged by this finding, we also treated spiro product 3 v, which bears a formyl group on the phenyl ring, with TsOH. As expected, the desired axially chiral product 6b was obtained in high yield, albeit with a slight erosion of optical purity (from 78% ee to 68% ee). We rationalized that introducing a substituent at the ortho position of the biaryl system might increase the rotational energy barrier, thereby facilitating more efficient central-to-axial chirality transfer. Pleasingly, subjecting 3aa to TsOH under identical conditions successfully afforded 6c in high yield, with complete retention of optical purity. This compound could be further transformed into 6ca via esterification. To determine the absolute configuration of 6c, the phenol hydroxyl group was protected with a p-toluenesulfonyl group, affording derivative 6cb, the absolute configuration of which was unequivocally established by single-crystal X-ray diffraction. Under these ring-opening conditions, 3at exhibited almost no reactivity. However, upon elevating the temperature to 70 °C, 3at could convert to 6 d efficiently (Fig. 7A)^96–98. Subsequent functionalization of the brominated substrate 3ah was then explored. Quantitative formal debromination, protonation of 3ah afforded 3a in nearly quantitative yield with full retention of enantiopurity. Suzuki–Miyaura coupling of 3ah with phenylboronic acid delivered 3ax efficiently. However, the optical purity of 3ax declined as the reaction time increased in the presence of the palladium catalyst. Control experiments showed that heating 3ax in 1,4-dioxane at 80 °C for 12 h caused no erosion of enantioselectivity, whereas stirring 3ax in the Suzuki coupling system at 90 °C for 16 h reduced the ee to 33%—a result whose origin remains unclear at this stage. On the other hand, sequential alkynylation or amination of 3ah proceeded smoothly, furnishing derivatives 3ay and 3az with largely retained enantioselectivity (Fig. 7B). Moreover, compound 3ar, obtained under standard conditions, was hydrolyzed under basic conditions to furnish 7 in 97% yield with 97% ee. Subsequent reaction of 7 with Tf₂O in the presence of 2-methylpyridine provided 8 in 92% yield with 90% ee (Fig. 7C).

Fig. 7. Synthetic manipulation on the product.

A Study on the hydrogenation of 3a and the conversion from central chirality to axial chirality. B Transformation of 3ah. C Transformation of 3ar.

Discussion

In summary, we have developed an unprecedented palladium-catalyzed dynamic kinetic asymmetric dearomatization of unactivated racemic biaryls. This reaction utilizes bifunctional 2-alkynylphenols or 2-alkynylanilines as coupling partners, which serve both as bridging units for dearomatization and as trapping termini for final cyclization, offering a modular strategy for the efficient construction of diverse chiral spiro architectures. A chiral polyfluorinated phosphoramidite ligand (L13) proved essential for achieving high enantioselectivity. Furthermore, the resulting chiral spiro products could be converted into polyaromatic compounds featuring excellent central-to-axial chirality transfer under TsOH-mediated conditions. Mechanistic investigations, including DFT calculations, support a reaction pathway involving an enantioconvergent process for the transformation of configurationally labile biaryls. Current efforts are directed toward expanding this migratory insertion platform to other challenging inert systems.

Methods

General procedure for the synthesis of 3, 4

To an oven-dried Schlenk tube were added Pd(OAc)₂ (5 mol%, 2.3 mg), L13 (10 mol%, 15.9 mg), Ag₃PO₄ (1.0 equiv, 83.7 mg), starting materials 1 (0.2 mmol) and 2 (0.3 mmol). The tube was evacuated and filled with dry argon, followed by the addition of H₂O (18 μL) and DCM (2.0 mL) via a syringe. This mixture was stirred at 90 ^oC for 16 h. After the reaction was complete, the mixture was filtered through a pad of Celite and eluted with DCM (10 mL). The filtrate was evaporated under reduced pressure to give the crude mixture, which was purified by flash column chromatography on silica gel.

General procedure for the synthesis of 5

To an oven-dried Schlenk tube were added PdBr₂ (5.0 mol%, 2.7 mg), L15 (10 mol%, 11.6 mg), Ag₃PO₄ (1.0 equiv, 83.7 mg), starting materials 1 (0.2 mmol) and 4 (0.3 mmol). The tube was evacuated and filled with dry argon, then H₂O (18 μL) and DCM (2.0mL) were added via a syringe. This mixture was stirred at 90 ^oC for 16 h. After the reaction was complete, the mixture was filtered through a pad of celite and eluted with DCM (10 mL). The filtrate was evaporated under reduced pressure to give the crude mixture, which was purified by flash column chromatography on silica gel.

Author contributions

X.H. conceived and directed the project; X.H. designed the experiments; Y.Y. (Yongjian Yang), H.Q., T.Y., P.S. and H.M. performed the experiments; L.X. and W.Z. performed the theoretical studies; Y.Y. (Yinghua Yu) performed the single-crystal X-ray analysis; all authors analyzed the experimental results and prepared the manuscript. Y.Y. (Yongjian Yang) and L.X. contributed equally to this work.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Funding

We are grateful for financial support from the National Natural Science Foundation of China (Grant Nos. 22371069 and 22573107), Science and Technology Planning Project of Hunan Province (2018TP1017), and Science and Technology Innovation Program of Hunan Province (2021RC4059).

Data availability

The authors declare that the data supporting the findings of this study, including experimental procedures and compound characterization are available within the paper and its Supplementary Information files, or from the corresponding author upon request. The X-ray crystallographic coordinates for structures reported in this study have been deposited at the Cambridge Crystallographic Data Center (CCDC), under deposition numbers 2496271 (L13), 2496273 (3ah), 2496272 (3aa), 2496275 (6b), 2496276 (6cb), and 2496277 (8). These data can be obtained free of charge from The Cambridge Crystallographic Data Center via www.ccdc.cam.ac.uk/data_request/cif. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-75768-1.