Introduction

Helicenes and helicenoids are a class of helical molecules in which steric repulsion between terminal aromatic rings induces a twisted, helical framework1. This distortion of the π-conjugated system confers unique chiroptical and electrical properties, such as optical rotation, electronic circular dichroism (CD), and circularly polarized luminescence (CPL)2,3,4,5. Due to these distinctive features, the precise synthesis of enantiomerically enriched helicenes with diverse architectures has attracted considerable interest in material science, and chiral catalysis6,7,8,9,10,11. Compared to the flourishing progress in central chirality construction, synthetic strategies for helical chirality remain relatively limited. Among the available synthetic strategies, transition metal catalyzed asymmetric transformations, including [2 + 2 + 2] cycloisomerization, hydroarylation, olefin metathesis and biaryl coupling, have significantly advanced the construction of chiral helicenes and helicenoids12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39. For example, in 2009, Tanaka and coworkers utilized a rhodium(I)/chiral bis-phosphine-catalyzed asymmetric intermolecular double [2 + 2 + 2] cycloisomerization to assemble a [9]helical fluorenone framework, albeit with moderate yield and enantioselectivity40. Later, this approach enabled efficient construction of [7]helical fluorenones (Fig. 1A)41. Alternatively, the Oestreich’s group attempted to construct [7]helical fluorenones via a palladium-catalyzed carbonylative reaction of optically pure axially chiral biaryls (Fig. 1B)42. However the axial chiraltiy transferred to helical chirality did not work well as a result of racemization at high reaction temperature. The design of innovative and general enantioselective synthetic routes to access new families of chiral substituted helical fluorenones and their structurally diverse derivatives remains a significant and ongoing challenge.

Fig. 1: Background information about asymmetric synthesis of helical fluorenones.
figure 1

A Rhodium(I)-catalyzed enantioselective intermolecular double [2 + 2 + 2] cycloaddition. B Palladium(II)-catalyzed carbonylative reaction using optically pure binaphthol sulfonate as the starting materials. C Kinetic resolution of helical fluorenones and synthesis of helical phenanthrenones via enantioselective homologation.

Full size image

The kinetic resolution strategy offers a complementary approach to simultaneously access enantiomerically pure starting materials and their derivatives. Moreover, carbonyl homologation using α-diazo compounds43,44 represents a powerful tool for precision single-atom skeletal editing in organic synthesis. We envisioned that integrating this homologation process into the kinetic resolution of helical fluorenones could concurrently enable the construction of structurally diverse helical phenanthrenones (Fig. 1C). Building on our prior work in enantioselective carbonyl homologation and others45,46,47,48,49,50,51,52,53,54, we identified several key challenges in adapting this transformation to helical fluorenones: (1) Catalyst design: a stronge Lewis acid catalyst paired with a tailored chiral ligand is essential to promote ketone homologation, leveraging ligand-accelerated cataysis55. (2) Enantiomer discrimiantion: effective resolution requires precise discrimination of the two up-and-down terminal aromatic rings distal to the metal center. (3) Stereoselectivity: the helical phenanthrenone product incorporate both helical and central chiralities, complicating diastereo- and enantiocontrol due to the symmetric, unbiased ketone substitution pattern. In addition, the reaction become further challenging with regioselectivty for un-symmetric ketones. Herein, we report a mild and efficient enantioselective homologation of racemic helical fluorenones by a chiral scandium(III) complex (Fig. 1C). Unlike traditional resolution methods, this strategy eliminates the need for chiral auxiliaries or chromatographic separation. A broad range of optically pure symmetric and unsymmetric [6/7]helical fluorenones and their ring-expanded [5/6/7]helical phenanthrenones, as well as other modified helical compounds were synthesized in high yield and high optical activity. Through stereodivergent synthesis, we accessed both enantiomers of [7]helical fluorenone and all four stereoisomers of [7]helical phenanthrenone. These compounds enable systemically investigation of photophysical and chiroptical properties, revealing strong correlation between structures and |gabs| values.

Results

Investigation of reaction conditions

Our study began with an evaluation of the kinetic resolution/skeletal editing of [7]helical fluorenone A1 with methyl α-diazo ester B1’ by the use of various chiral scandium(III) catalysts (for details see Supplementary Information Table S1–8). The homologation reaction was highly dependent on the catalyst components, occurring only in the presence of particular N,N’-dioxide ligands and with Sc(OTf)3 as the metal precursor. For ligand skeletons, L-pipecolic acid derived ones were favorable in terms of reactivity and enantioselectivity, beating other amino acid derived analogues (for details see Supplementary Information Table S2). As shown in Table 1, when L-pipecolic acid-based N,N’-dioxides were evaluated, it was found that the reactivity is also sensitive to the ligand structures with variation at amide substitutions. Ligand L3-PiPr2 containing steric hindered 2,6-diisopropyl aniline substituents hampered the reaction, leading to no homologation reaction and resolution (Table 1, entry 3 vs entries 1 and 2). Thus, consequent ligand modification focused on introducing substitution at the para-position of 2,6-dimethylaniline to expand the chiral pocket for discrimination of terminal aryls (Table 1, entries 4-6). Installing para-diphenylmethanyl substituents (ligand NO-1) exhibited minimal impact on the reaction performance. Notably, the introduction of conjugate naphthyl groups significantly altered the enantioselectivity and diastereoselectivity of the kinetic resolution/skeletal editing process (Table 1, entries 5 and 6). When employing ligand NO-3, the [7]helical phenanthrenone derivative C1’ was isolated in 31% yield with enantioselectivity being enhanced to 91% ee of the major diastereomer (7.0:1 dr), while the starting material A1 was recovered in 32% e.e. (Table 1, entry 6).

Table 1 Reaction development and standard conditions
Full size table

Through reduce of reaction concentration and adjust the substrates ratio (Table 1, entry 7), the conversion increased to 48% along with slightly higher ee value for helical fluorenone A1 (57% e.e.). Next, we examined the influence of the ester group of α-diazo compounds to further improve the results (for more details see Supplementary Information Table S7). When iso-butyl diazoester B1 was used, the corresponding [7]helical phenanthrenone derivative C1 was isolated in 57% yield, 5.2:1 dr and up to 95% e.e., meanwhile the helical fluorenone A1 was recovered in 43% yield and 87% e.e. (Table 1, entry 8). Decreased reaction temperature was beneficial to diastereoselectivity but the conversion was reduced (Table 1, entry 9). Through systematic investigation of solvents and temperature (for details see Supplementary Information Tables S4 and S6), we ultimately discovered that employing 1,1,2,2-TCE as solvent at 0 °C significantly improved the ee value of the product C1 to 98% without compromising conversion efficiency, while enabling 51% recovery of helical fluorenone A1 with 80% e.e. (Table 1 entry 10). To solve the poor solubility of Sc(OTf)₃ in 1,1,2,2-TCE, THF was employed as pro-solvent for chiral catalyst preparation. This pre-treatment significantly enhanced the catalytic activity (Table 1, entries 11 and 12). Ultimately, kinetic resolution through homologation after controlled reaction time gave raise to two kinds of helical chiral compounds in considerable yield and excellent enantioselectivity (entry 11; C1/98% e.e., and A1/94% e.e.). Following the similar procedure using other chiral ligands, such as BINOL and BOX, no desired transformation occurred (Table 1, entry 13), manifesting obvious ligand acceleration55.

Substrates scope of helical fluorenones

With the optimized reaction conditions in hand, the substrate scope of helical fluorenones was evaluated (Fig. 2). A series of optically enriched [7]helical phenanthrenone derivatives (C2−C17) and substituted [7]helical fluorenones (A2−A17) were available through homologation-involved kinetic resolution. On the whole, the position of substituents on the aryl ring of [7]helical fluorenones exhibited minimal influence on the outcome of the reaction, except for those with substituents at the 4,4’-position (A4, A9 and A16). For these three cases, it is not easy to leverage the stereoselectivity and yield, and these products were obtained in high diastereo- and enantioselectivity at the cost of conversion or optical activity of the recovered substrates. The homologation activity is closely related to the electronic effect of the substituents. Substrates (A2−A6, A10 and A12−A14) with electron-donating alkyl groups, regardless positions and steric hindrance, exhibited higher reactivity, requiring shorter reaction times (4−5 h) than helical fluorenones (A7−A9, A11 and A17) with electron-withdrawing aryl groups (8−12 h). These reactions delivered the desired [7]helical phenanthrenone derivatives (C2−C17) in 42–50% yield, 4.9:1 to 13.7:1 dr and 91–99% e.e. for all diastereomers. The corresponding [7]helical fluorenones could be recovered in 38–47% yield with 80–97% e.e. Racemic phenyl-substituted A15 underwent the reaction in a stereoselective manner to yield C15 in 45% yield, 11.2:1 d.r. and 97% ee after a longer time, and kinetic resolution lead to enantiomerically enrich A15 (90% e.e.) in 45% yield. Heterocycle-bearing [7]helical fluorenones, such as thiophene-fused A18 and furan-fused A19 exhibit higher reactivity and resolution efficiency within 3 h, but the excellent results maintained even after 24 h. The process could also be applied to naphthalene-fused [9]helical fluorenone A20 although its behavior was similar to 4,4’-substituted ones. The related [9]helical phenanthrenone C20 was isolated in 41% yield, 8.4:1 d.r. with 97% e.e., accompanied by recovering A20 in 49% yield with 77% e.e.

Fig. 2: Substrate scope of helical fluorenones.
figure 2

All reactions were carried out with helical fluorenone A (0.05 mmol), diazo ester B1 (1.25 equiv.), Sc(OTf)3/NO-3 (1.3:1, 10 mol%) in 1,1,2,2-TCE (0.025 M) under air atmosphere. Isolated yields after column chromatography separation. The d.r. value was determined by 1H NMR spectroscopy and the e.e. value was determined by HPLC.

Full size image

Through X-ray crystallographic analysis, the recovered [7]helical fluorenone A18 was determined to be M-configuration (Fig. 2). The other recovered helical fluorenones were also assigned as the same configuration based on circular dichroism (CD) spectroscopy analysis. Oppositely, the P-configured isomers prefer to perform the ring-expansion under the catalysis. The relative configuration of the major diastereomer of the racemic [7]helical phenanthrenone C6 is assigned as rel-(P, S) or rel-(M, R) based on X-ray crystallographic analysis. Accordingly, the absolute configuration of asymmetric catalytic product C6 was assigned as (P, S). The configuration of the other major isomers of helical phenanthrenones following the same arrangement from the corresponding CD spectroscopy.

The catalytic system also enables efficient kinetic resolution of unsymmetric [6/7]helical fluorenones (Fig. 3), manifesting good regio-, diastereo-, and enantioselectivity. Similarly, (M)-fluoreones (A21–A25) were recovered as the less reactive isomers (36–54% yields, 67–98% e.e.). (P)-fluoreones (A21–A25) participated in homologation to give the ring-expanded [6/7]helical phenanthrenones (C21–C25) in which small substituent (Ar2) shifted to generate the quaternary carbon center with good yields and ee values (43–50% yields, 93–95% e.e.). In comparison, (P)-[6]helical fluorenones (A21–A23) demonstrated high reactivity to achieve good yields within 3 h, accompanied by around 5% yield of isomers (C21’–C23’). The [6]helical fluorenone A24 and [7]helical fluorenone A25 exhibited diminished reactivity and longer reaction time was required. And the former one suffered ring-expansion of (M)-isomer to yield (M, S)-C24’ in 14% isolated yield and 96% e.e., which generated from the migration of fused-aryl group based on X-ray crystal analysis. The arrangement of the major product C25 was assigned as (P, S) concluded from the X-ray crystal of the related racemate as well as CD spectra. In addition, [5]helicene fluorenones A26 and A27 underwent complete conversion, affording the ring-expanded product C26 with high enantioselectivity and diastereoselectivity (89% yield, 18/1 d.r., 93% e.e.) at 35 °C, and two diastereomers of C27 (1.3:1 d.r.) with comparable high enantioselectivity at 35 °C, respectively (for details see Supplementary Information Table S9).

Fig. 3: Substrate scope of unsymmetric helical ketones and others.
figure 3

Unless otherwise noted, the reaction condition is as same as in Fig. 2.

Full size image

Substrates scope of α-Diazo esters

To further explore the substitution variety of [7]helical phenanthrenones, a number of α-diazo esters with different substituents were tested (Fig. 4). It was obvious that α-diazo esters bearing electron-donating substituted benzyl groups participated in the reaction to get satisfied yield within shorter reaction time (less than 5 h), in comparison with these bearing electron-withdrawing groups (9–54 h). Substitutions at meta- and para-positions were all tolerable well, and different substituted [7]helical phenanthrenones (D1–D25) could be isolated in excellent enantioselectivity (95–99% e.e.). The para-methoxy substituted ones were more reactivity, and over-conversion resulted in reduced d.r. value (D8, D21 and D25) and higher yield even within short reaction time, however, the diastereoselectivity of the others could achieve 6.0:1 to 15:1 at 40–50% yield. Noteworthily, 3-nitro-substitued D18 could be obtained in 40% yield after longer reaction time. For cyanide and nitro-substituted ones (D13, D17, and D18), the reactivity drastically decreased, minimizing further occurrence of homologation even after prolonged reaction times. Both 2-naphthyl and 1-naphthyl substituted α-diazo esters could be edited into the helical structures (D26 and D27), but the latter one required longer reaction time to achieve acceptable yield, with extremely high diastereo- and enantioselectivity (up to 19:1 d.r., and 98–99% e.e.). The benzofuran-containing D28 and cinnamyl-containing D30 were isolated in excellent results. In addition, indole and pyridine-substituted α-diazoesters were also tested, unfortunately, there is lack of analytical methods to determine the stereoselectivity of the products. In these reactions, the unreacted [7]helical fluorenone A1 could be isolated in 35–52% yields, meanwhile the enantioselectivity dropped along with higher recover yield, confirming a kinetic resolution process. Alkyl-substituted diazo esters, such as phenethyl and stearyl derivatives, indeed exhibited attenuated reactivity. Under standard catalyst after 24 h, these substrates afford the products (D31 and D32) in moderate yields with good enantioselectivity. In contrast, diazoesters bearing allyl, alkenyl, hydrogen, or phenyl substituents (C33–C36) demonstrate markedly diminished reactivity.

Fig. 4: Substrate scope of α-diazo esters.
figure 4

Unless otherwise noted, the reactions were carried out with helical fluorenone A1 (0.05 mmol), diazo compound B (1.25 equiv.), Sc(OTf)3/NO-3 (1.3:1, 10 mol%) in 1,1,2,2-TCE (0.025 M) under air atmosphere. Isolated yields after column chromatography separation. The d.r. value was determined by 1H NMR spectroscopy and the e.e. was determined by HPLC.

Full size image

Stereodivergent synthesis and synthetic applications

The potential of this method for the skeletal editing-based kinetic resolution of [7]helical fluorenones was further demonstrated by gram-scale synthesis. As shown in Fig. 5A, the stereodivergent synthesis was carried out with racemic [7]helical fluorenone A1 (5.0 mmol) and α-diazo ester B1 (6.25 mmol) under the standard reaction conditions catalyzed by Sc(III)/NO-3 or Sc(III)/ent-NO-3. Thus, the [7]helical phenanthrenone product (P, S)-C1 or (M, R)-C1 could be isolated in comparable yield and stereoselectivity, at the same time, [7]helical fluorenone (M)-A1 or (P)-A1 could be obtained separately. Scale-up reaction performed well under the standard condition, and even reduced reaction time (3 h) was tolerable. As shown in Figs. 2 and 3, in all cases, the minor diastereomers could also be isolated in excellent enantioselectivity, manifested that chiral quaternary centers could be well controlled by the catalyst, which give chance for stereodivergent synthesis. If the isolated optically active M-A1 and P-A1 were resubjected into the original catalytic systems, after full transformations, the corresponding (S, M)-C1 and (R, P)-C1 could be isolated in up to >19/1 d.r. with 97% ee. Therefore, stereodivergent synthesis of the four isomers of [7]helical phenanthrenone bearing central and helical chirality could be readily achieved.

Fig. 5: Gram-scale synthesis, stereodivergent synthesis and synthetic applications.
figure 5

A Gram-scale synthesis and stereodivergent synthesis all stereoisomers of A1 and C1. B Further transformations of the (M)-A1 and (P,S)-C29. Isolated yields after column chromatography separation.

Full size image

The synthetic transformations of the [7]helical fluorenone M-A1 enabled the preparation of a series of [7]helical compounds (Fig. 5B). Reduction with NaBH4 gave raise to (M)-[7]helical fluorenol E1 in nearly quantitative yield. Both methyl and phenyl substituted [7]helical fluorenol (M)-E2 and (M)-E3 could be isolated in good yield after the treatment with Grignard reagents. The former E2 has been used for synthesis of helically chiral carbenium ion, acting as carbon Lewis acid catalyst42. When using Et3SiH/BF3·OEt2 reducing system, the carbonyl group was directly reduced to a methylene group, delivering (M)-[7]helical fluorene E4 in 78% yield, accompanied by un-identified by-products. The corresponding O-methyloxime (M)-E5 and hydrazone (M)-E6 could be quantitatively isolated after mixing helical fluorenone with NH2OMe·or hydrazine hydrate in pyridine. The hydrazone (M)-E6 could be further converted to a diazo-containing helical compound (M)-E7 using manganese dioxide as the oxidant and magnesium sulfate as a dehydrating agent. [7]Helical fluorene fused spiro cyclic derivative (M)-E8 could be obtained from the coupling with 1-bromo-2-benzobenzene. By employing BF3·OEt2 as the catalyst, skeletal editing of (M)-A1 with TMSCHN2 led to efficient synthesis of [7]helicene (M)-E9 after acylation with isopropyl chloroformate. Moreover, upon undergoing straightforward decarbonylation followed by acylation protection, the ring-expanded product (P, S)-D29 can be converted to the all-carbon-conjugated [7]helicene derivative (P)-E10 in good yield. In all these transformations, the helical chirality could be maintained.

Photophysical chiroptical properties

As illustrated in Fig. 6A, [7]helical fluorenone (M)-A1 and its enantiomer (P)-A1 display a mirror image of the spectral shapes of circular dichroism in dichloromethane. The robust Cotton effect in the high-energy band from 280 to 370 nm can be partly attributed to a mixture of n→π* and π → π* transition. There is weak and broad Cotton effect appeared in the low-energy band from 420 to 520 nm. The enantiomeric pair of [7]helical phenanthrenone (P, S)-C1 and (M, R)-C1 also exhibited mirror-symmetric CD spectra, and the diastereomer (P, R)-C1 and (P, S)-C1 had similar Cotton effect at 350 nm. The ring-expanded phenanthrenones showed blue shift of the absorption compared with fluorenone in the high-energy band, and the Cotton effect in the low-energy band disappeared. The dissymmetry factors (gabs), calculated with the relationship gabs = Δε/ε, revealed distinct wavelength-dependent profiles. The maximum |gabs| of [7]helical fluorenone M-A1 and [7]helical phenanthrenone (P, S)-C1 reached −7.7 × 10−3 (at 362 nm) and +1.4 × 10−2 (at 349 nm), respectively. The ring expansion and introduction of a chiral quaternary carbon center nearly doubled the asymmetric absorption factor compared to the [7]helical fluorenone A1.

Fig. 6: Photophysical chiroptical properties studies.
figure 6

A Circular dichroism (CD) spectra of A1 and C1. B gabs values plot of A1 and C1. C, D) TD-DFT calculated parameters of A1 and C1 at the PBE0/def2-TZVP theory level. μ: transition electric dipole moments. m: transition magnetic dipole moments. θ: angle between m and μ. E, F Comparison between theoretical (gcal) and experimental (gexp) dissymmetry factors.

Full size image

To elucidate the origin of the enhanced |gabs| values, time-dependent density functional theory (TD-DFT) calculations were conducted at the PBE0/def2-TZVP level56,57,58,59. For each enantiomer of A1, the transition corresponding to the largest |gabs| was identified as S₀ → S₅ transition, whereas for each pair of enantiomers of C1, the dominant transition shifted to S₀ → S₄ transition. Theoretical analysis defines the dissymmetry factor of small organic molecules is usually simplified as gabs ≈ 4cosθ | m | /|μ | . The gabs value exhibits a correlation with both the |m | /|μ| ratio and the angle θ between m and μ, where m and μ represent the magnetic and electric transition dipole moments, respectively. Computational analysis of the [7]helical fluorenone A1 and [7]helical phenanthrenone C1 quantified these parameters which were shown in Fig. 6C. The enhancement in |gabs| stems from two key factors: a higher |m | /|μ| ratio in [7]helical phenanthrenone C1 system compared to [7]helical fluorenone A1 as well as decreased θ, leading to stronger alignment of μ and m (Fig. 6D). These computational results align closely with experimental gabs measurements, confirming that structural modifications amplify chiroptical responses by synergistically enhancing magnetic transition contributions and dipole alignment.

Systematic experimental and theoretical investigations of absorptive dissymmetry factor across the helical fluorenone series (M-A1 to M-A20) revealed significant substituent-dependent modulation of chiroptical properties. Notably, positional isomerism markedly influenced the |gabs | . The installation of methyl group from the 2- to 6-position (M-A2 → M-A6) reduced |gabs| value from 8.3 × 10−3 to 5.2 × 10−3 (Fig. 6E, black labels). Next, by fixing the substituents at 4,4’-positions, it revealed that the steric effects substantially impacted chiroptical responses. Increasing the bulk of substituent from hydrogen (M-A1, |gabs | = 7.7 × 10−3) to phenyl (M-A15, |gabs | = 1.1 × 10−2) produced progressive enhancement of dissymmetry factors (Fig. 6F). In these examples, the theoretical gcal and gabs factors exhibited consistent trends with variations in peripheral substituents (Fig. 6E, F, red labels). In addition, enhanced conjugation substitutions significantly improved chiroptical performance, such as 3-phenyl substituted M-A15 and [9]-helical fluorenone A20 (gabs as 9.9 × 10−3 at 366 nm). This comprehensive study shows that strategic modification of peripheral substituents on helical fluorenones enables precise tuning of chiroptical response magnitude through synergistic control of electronic environment and steric demands. Furthermore, the circularly polarized luminescence (CPL) spectra of some products were tested (for details see Supplementary Information Figs. S103–S109). A mirror-image relationship was observed in of (P, S)-C1 and (M, R)-C1 in dichloromethane, with a luminescence dissymmetry factor |g1um| of 3.6 × 10−4 at 500 nm. (P, R)-C1 and (M, S)-C1 also exhibited mirror-imaged CPL spectra, in which the CPL emission band was red-shifted to approximately 505 nm, and the |g1um| value was determined to be 3.8 × 10−4. The inversion of stereocenters scarcely altered the CPL behavior, indicating that the results are primarily governed by the direction of the helical skeleton. Compound (P, S)-C19 exhibited the strongest CPL emission among the series, with its luminescence dissymmetry factor |g1um| measured at 1.1 × 10−3 at 560 nm. The helicenone substrate M-A10 displayed CPL in DCM solution at 540 nm with a |g1um| value of 1.2 × 10−3. These CD and CPL results clearly demonstrate that optically enriched [7]helical fluorenones and ring-expanded [7]helical phenanthrenones maintain their chirality in both the ground and excited states.

Discussion

In summary, we have achieved an enantioselective homologation of racemic helical fluorenones catalyzed by a scandium complex with a meticulously designed N,N’-dioxide ligand under mild reaction condition. This process provided access to a series of ring-expanded helical phenanthrenones with helical and central chiral elements in high diastereo- and enantioselectivity, meanwhile less reactive helical fluorenone enantiomers were efficiently recovered with high optical activity under kinetic resolution. For unsymmetric helical fluorenones, the aryl migration during the homologative ring expansion process also demonstrated interesting regioselectivity. Specifically, with P-helical fluorenones, the smaller aryl group preferentially migrates, whereas the opposite preference was observed conversely. Further transformation of helical fluorenone gave rise to several functionalized helical compounds. Moreover, the photophysical chiroptical properties of these helical compounds were systematically investigated, specifically examining the regulatory effects of substituents on the electronic circular dichroism of these helical compounds. Through synergistic integration of theoretical calculations and experimental data, we elucidated the mechanism underlying the enhanced electronic circular dichroism response in penanthrenones enabled by homologation-driven skeletal editing. This methodology not only provides an efficient approach for synthesizing optically pure helical fluorenones and helical phenanthrenones but also offers extensive experimental and theoretical insights for designing molecular architectures with enhanced electronic circular dichroism responses. Further application of the catalysts system for other transformations and explore the usage of these helical compounds are undergoing.

Methods

General procedure for the kinetic resolution of helical fluorenones

To a dry test tube under nitrogen atmosphere were added Sc(OTf)3 (3.2 mg, 0.0065 mmol, 13 mol%), ligand NO-3 (5.4 mg, 0.005 mmol, 10 mol%) in THF (1.0 mL), and the mixture was stirred at 35 °C for 20 min. Substrate A (19 mg, 0.05 mmol, 1.0 equiv.) was added to the test tube after THF was evaporated. Then the resulting mixture was stirred at 35 °C for 20 min, after cooling to 0 °C substrate B (18 μL, 0.0625 mmol, 1.5 equiv.) was added. The reaction solution was stirred under such conditions. After the ee value of spiral fluorenone A was increased to more than 90% (detected by chiral HPLC), the reaction solution was filtered by a short silica gel column to quench the reaction, and the solution was concentrated under reduced pressure. The crude residue was subjected to column chromatography on silica gel (2:1 v/v Pet:DCM as the eluent) to afford the product (P, S)-C and M-A.