Enantioselective C–H Activation for the Construction of Chiral Tertiary Fluorides via Migrative Carbofluorination

Abstract

Construction of α-chiral centers via enantioselective C(sp³)−H activation of the isobutyric acid scaffold has emerged as a new chiral technology. However, this approach has not been compatible with the presence of α-fluorine which prevented the access to fluorine-bearing α-chiral centers which are highly valuable in drug discovery. Here, we report a palladium-catalyzed asymmetric β-C(sp³)−H activation of the gem-dimethyl motif, followed by a dyotropic rearrangement that installs fluoro and aryl groups at the α- and β-positions, respectively. Using a bifunctional chiral monoprotected amino sulfonamide (MPASA) ligand, this method affords chiral tertiary fluorides in high reactivity and enantioselectivity, enabling access to desirable motifs in drug discovery.

Graphical Abstract

Pd-catalyzed enantioselective C(sp³)−H activation reactions of weakly coordinating substrates have emerged as a promising chiral technology for constructing α-, and β-chiral centers.¹ Recently, our group and Jiao’s group have developed the desymmetrizing β-C(sp³)−H arylation, olefination, oxidation, and fluorination of amides and lactams to construct α-chiral centers by developing a series of chiral bifunctional ligands.^2,3 In this endeavor, the incompatibility with α-halides and heteroatoms poses a significant limitation for constructing corresponding highly functionalized chiral centers that are broadly useful in organic synthesis, especially the α-fluorides.^4,5 The importance of chiral tertiary α-fluoro carbonyl compounds has drawn extensive attention which primarily focused the following two approaches: 1) asymmetric α-fluorination of carbonyl compounds with nucleophilic or electrophilic fluorinating reagents, 2) asymmetric transformation of α-fluorinated substrates.⁶ Evidently, the development of alternative methods for constructing chiral tertiary fluorides through enantioselective C(sp³)−H activation method remains highly desirable in synthesis and drug discovery (Figure 1A).⁷

Figure 1.

Enantioselective Pd-Catalyzed C(sp³)−H Functionalization of Native Amides

While our efforts to develop new chiral ligands for desymmetrizing C(sp³)−H activation of isobutyric acid scaffold bearing α-fluoride has not been fruitful so far, an interesting dyotropic rearrangement initiated by β-C(sp³)−H activation of Weinreb amide discovered by the Zhu lab caught our attention⁸ (Figure 1B). We envision such chemical process could be potentially rendered enantioselective to construct chiral tertiary fluorides if a chiral ligand can be identified to control the stereochemistry of the C(sp³)−H activation step to form chiral intermediate A. Subsequently, A is oxidized to Pd^IV by Selectfluor to afford intermediate B, which undergoes a dyotropic rearrangement process. A backside attack by fluorine then induces inversion of configuration, and the final reductive elimination delivers chiral tertiary fluorides product 2 (Figure 1C).⁹ However, the realization of this proposed enantioselective reaction is expected to meet with several challenges. First, as reported by Zhu, this reaction proceeds efficiently in the absence of ligand, and the resulting strong background reactivity presents a considerable challenge for achieving stereocontrol.⁸ Second, achieving differentiation between the subtle stereochemical space of methyl and phenyl substituents remains a significant challenge.¹⁰ Third, undesired β-C(sp³)–H fluorination is potentially a competing pathway.¹¹ Herein we report the development of a palladium-catalyzed enantioselective migrative carbofluorination reaction, providing an alternative route for constructing chiral tertiary fluorides with up to 99:1 er. The combination of less reactive Pd(PPh₃)₂Cl₂ and the effective ligand acceleration with the bifunctional chiral monoprotected amino sulfonamide (MPASA) ligand is crucial to suppress the background reaction and achieve excellent enantioselectivity.

Considering the pivotal role of chiral ligands in stereocontrol of asymmetric C–H functionlizations,^{12, 13} we initiated our efforts by focusing on chiral ligands screening with 2-methyl-2-phenyl-1-(piperidin-1-yl)propan-1-one (3a) as the model substrate and Selectfluor as the fluorine source (Figure 2). The chiral bidentate mono-N-acetyl amino acid and amino O-methylhydroxamic acid (MPAA) ligand (L1, L2) provided the desired product with 12–21% yields and modest enantioselectivities along with trace amounts of the β-fluorinated product (4a’) (2–4% yields) (Figure 2, entries 1−2). Other chiral ligands (L3–L6) such as the mono-N-acetyl aminoethyl phenyl thioether (MPAThio), aminoethyl amine (MPAAM), amino quinoline (MPAQ) showed 6–17% yields and lower enantioselectivities (entries 3−6). Subsequently, the performance of the MPASA ligands (L7–L9) were investigated (see Supporting Information). It was found that installing an electron-withdrawing substituent on the aryl sulfonamide significantly improved the reactivity to 62% yield and the enantioselectivity to 82:18 er (entry 8). The replacement of the isopropyl group with other substituents on the side chain slightly decreased enantioselectivities (see Supporting Information). A slight improvement in enantioselectivity was obtained by replacing Ag₂CO₃ with Ag₃PO₄ (entry 9). We hypothesized that the acetate anion of Pd(OAc)₂ could potentially serve as a CMD base to promote the C–H activation process, thereby enhancing the background reactivity and diminishing stereoselectivity. As expected, the enantioselectivity was markedly improved to 94:8 er when Pd(PPh₃)₂Cl₂ was employed as the palladium source (entry 10). Consequently, employing Pd(PPh₃)₂Cl₂ together with Ag₃PO₄ proved optimal, delivering the desired product in 66% yield and 94.5:5.5 er.

Figure 2.

Reaction Condition Screening for enantioselective Migrative Carbofluorination. ^aReaction conditions: 3a (0.1 mmol), Selectfluor (0.2 mmol), Pd(OAc)₂ (10 mol%), Ligand (12 mol%), Ag₂CO₃ (0.5 equiv) HFIP (0.5 mL) at 50 °C for 48 hours. The yields were determined by ¹H NMR analysis of the crude product using CH₂Br₂ as the internal standard. The er were determined by chiral SFC. ^bL9 instead of L1. *Isolated yield.

With the optimal reaction conditions in hand, we first investigated the substrate scope of α-aryl groups for this enantioselective C(sp³)–H migrative carbofluorination reaction (Figure 3). Aryl groups bearing electron-donating substituents at para-positions were well tolerated, delivering the corresponding products (2a–c) in good yields and excellent enantioselectivities (95.5:4.5 to 96.5:3.5 er). Moreover, para-halogen substituents such as fluorine, chlorine, and bromine were compatible in our reaction conditions, affording the desired products (2d–f) in 61–78% yields with enantioselectivities ranging from 90.5:9.5 to 96.6:3.5 er. Aryl groups bearing ortho substituents also performed well in the reaction, forming the desired products (2g, 2h) in moderate yields and high enantioselectivities (96:4 to 96.5:3.5 er). Interestingly, the meta bromine-substituted aryl groups (2i) resulted in a significant drop in reactivity while enantioselectivity remains high. Reaction of the amide derived from flurbiprofen also delivered the desired carbofluorination product (2j) with moderate yield and excellent enantioselectivity (95.5:4.5 er). Ortho-, para-dihalogenated aryl substituted substrates reacted smoothly with good yields and excellent enantioselectivities (2k, 2l).

Figure 3.

Scope of Substrate. Reaction conditions: 3 (0.1 mmol), Selectfluor (0.2 mmol), Pd(Ph₃P)₂Cl₂ (10 mol%), L9 (12 mol%), Ag₃PO₄ (0.5 equiv) HFIP (0.5 mL) at 50 °C for 48 hours. Isolated yield. The er were determined by chiral SFC. Isolated yield.

Next, we examined the scope of the amine moiety of native amides (Figure 4). The amide (3b) from the noncyclic amine was found to react smoothly with good yield but much lower enantioselectivity. While the enantioselectivity can be restored using bulkier acyclic amide, the yield dropped significantly (see supporting information). Amide 3c from azepane was also well tolerated, providing the product (4c) in moderate yield and highest enantioselectivity (99:1 er). Amides derived from a wide range of 4-substituted piperidines including chlorine (3d), bromine (3e) and methoxy (3f) were all reactive, affording the desired products in moderate yields and good enantioselectivities. Keto and ester groups on the piperidines (4g–i) were also compatible, giving moderate yields and good enantioselectivities (94:6 to 95.5:4.5 er). In addition, di-substitution on the piperidine was also tolerated (4j and 4k). Notably, amide from Ts-protected piperazine was also a suitable substrate, providing corresponding product (4l) in 56% yield and 95:5 er. It is worth-noting that these amides derived from pieridine and piperazine can be reduced to medicinally relevant chiral saturated heterocycles. Finally, the absolute configuration of major enantiomer of the product R-4l was determined by single crystal X-ray diffraction, which provides strong evidence for the proposed stereomodel for this enantioselective C–H activation cascade reaction.

Figure. 4.

Substrate Scope. Reaction conditions: 3 (0.1 mmol), Selectfluor (0.2 mmol), Pd(Ph₃P)₂Cl₂ (10 mol%), L9 (12 mol%), Ag₃PO₄ (0.5 equiv) HFIP (0.5 mL) at 50 °C for 48 hours. Isolated yield. The er were determined by chiral SFC. Isolated yield.

The scalability of the developed protocol was also examined by conducting the reaction on 2.0 mmol scale, affording the desired product in 61% yield with 97:3 er (Figure 5A). To further elucidate the reaction pathway, mechanistic experiments were conducted. The absence of deuterium incorporation in the migrative carbofluorination product and recovered substrate suggests that the β-C(sp³)–H bond cleavage step could be irreversible (Figure 5B). Significant primary kinetic isotope effects were observed in both the parallel (pKIE = 2.9) and competitive (cKIE = 3.2) experiments of substrates 1l and d-1l (Figure 5C), consistent with the conclusion that enantioselective β-C(sp³)–H activation is involved in the turnover-limiting step.

Figure. 5.

Large-scale reaction and mechanism studies

In summary, we have developed an enantioselective method for the construction of chiral tertiary fluorides by merging asymmetric β-C(sp³)–H activation with dyotropic rearrangement. This approach fills a gap for current enantioselective C(sp³)–H activation reactions that are incompatible with α-fluorides.

Supplementary Material

The Supporting Information is available free of charge on the ACS Publications website. Experimental details, full characterization of new compounds including ¹H and ¹³C NMR spectra, HRMS data (PDF)