Synthesis of Spiro Carbocycles via Methyl β-C(sp³)–H Functionalization of Cyclic Aliphatic Acids

Abstract

Despite the superior innate reactivity of methyl C–H bonds over methylene C–H bonds in acyclic substrates, selective activation of methyl C–H bonds in carbocycles remains a significant challenge. Recent effective bifunctional pyridone ligands have consistently demonstrated selectivity towards methylene C–H bonds on various cyclic carboxylic acids. Selective functionalization of β-methyl C–H bonds of cyclic carboxylic acids could further expand diversity of carbocycles, especially as this reactivity could open a new route for the synthesis of spirocyclic scaffolds desirable in medicinal chemistry. Herein, we report a ligand-controlled β-C(sp³)–H methyl olefination and arylation that enables the efficient synthesis of spirocyclic lactones, spiro-3,4-dihydrocoumarins and spirocyclic ketones. Computational and deuterium incorporation studies suggest that MPAA (mono-N-protected amino acid) and MPAThio (mono-N-protected amino aryl thioether) ligands reverse the preference for methylene C–H activation observed with bidentate pyridone ligands by favoring the activation of primary C–H bonds. Furthermore, sequential functionalization of both methyl and methylene C–H bonds in cyclic acids were realized to significantly expand the structural diversity of carbocycles.

Graphical Abstract

1. Introduction

Cyclic aliphatic acids are ubiquitous in natural products and pharmaceutical agents.¹ Their site-selective functionalization represents a powerful strategy for the construction of structurally complex and biologically active molecules, underscoring their importance in drug discovery. Over the past decade, the development of MPAA and MPAThio ligands has enabled a variety of directed β-C(sp³)–H methyl activation reactions on acyclic aliphatic acids,² offering unprecedented synthetic disconnections. For instance, these ligands have facilitated β-C(sp³)–H olefination,³ carbonylation,⁴ arylation,⁵ acyloxylation,⁶ and lactonization⁷ while pyridone-acid ligand has been shown to promote hydroxylation⁸ (Scheme 1A). While methyl C–H bonds are intrinsically more reactive than methylene C–H bonds in acyclic substrates, the selective activation of β-methyl C–H bonds in cyclic frameworks remains a formidable challenge. In contrast, the recent development of bidentate pyridone ligands has enabled a series of methylene-selective β-C(sp³)–H functionalizations in cyclic carboxylic acids. These include pyridone–imine and pyridone–amine-promoted cascade β,γ-dehydrogenation/olefination,⁹ β,γ-dehydrogenation¹⁰ and γ-arylation facilitated by a pyridone–quinuclidine ligand¹¹, and [2+2] benzannulation enabled by a pyridone–amide system (Scheme 1B).¹² Although amino acid–derived ligands are generally inefficient for methylene C–H activation, two notable exceptions have been reported: β,γ-diarylation promoted by MPAThio ligands¹³ and γ-lactonization facilitated by MPAA ligands¹⁴. Despite these advances, activation of methyl C–H bonds remains a significant challenge. The selective functionalization of β-methyl C–H bonds in cyclic carboxylic acids could significantly expand the structural diversity of carbocycles by providing a new synthetic route to spirocyclic scaffolds, which are considered privileged structures in drug discovery due to their rigid three-dimensional architecture that enhances molecular complexity, improves pharmacokinetic properties, and contributes to target selectivity.¹⁵

Scheme 1.

Ligand-Controlled Site-selective Functionalization of Carboxylic Acids via β-C(sp³)–H Activation

In this work, we report a ligand-controlled selective β-C(sp³)–H methyl olefination and arylation, enabling the efficient synthesis of diverse spirocyclic lactones, spiro-3,4-dihydrocoumarins, and spirocyclic ketones. (Scheme 1B). Computational and deuterium incorporation revealed that MPAA and MPAThio ligands showed preferential palladation at the methyl site. In addition, we developed sequential strategies to access more complex and highly functionalized carbocycles via sequential ligand-controlled C(sp³)–H methyl and methylene functionalization (Scheme 1C).

2. Results and Discussion

Our studies on β-C(sp³)–H methyl olefination of cyclic aliphatic acids began with cyclopentane carboxylic acid 1a and benzyl acrylate 2a as the model acid substrate and olefin coupling partner, respectively. Based on our previously reported conditions for racemic and enantioselective β,γ-dehydrogenation/olefination and β,γ-dehydrogenation of carboxylic acids and tosyl-protected amides, we introduced acetonitrile (MeCN) as a co-solvent, which played a crucial role in enhancing reaction efficiency.^{9, 10, 16} A 10:1 mixture of HFIP and MeCN was employed as the solvent system, with KF serving as the base in our initial reaction conditions. Interestingly, the target lactone product was observed in 14% yield even in the absence of a ligand. Next, we focused on identifying ligands that could significantly improve this reaction (Table 1). Guided by prior successes in ligand-enabled C(sp³)–H activation, we evaluated both pyridine and monodentate pyridone ligands previously developed by our group.¹⁷ While these ligands did improve the yield of the desired lactone products, the overall yields remained modest (22% and 25%, respectively). As such, a series of MPAA (L3, L4, L5, L7) and MPAAM (L6) ligands, known to facilitate β-lactonization of free carboxylic acids were evaluated. Among them, the Ac-Phe-OH (L7) ligand improving the yield to 76 %. Using ligand L7, we conducted an oxidant screening, which identified silver carbonate as the optimal oxidant (Table S3). The bidentate ligands L8-L10, which had previously proven effective in weakly coordinating group–directed C(sp³)–H activations, were also examined. The MPAThio ligand, earlier developed for the C(sp³)−H olefination³ and carbonylation⁴ of free carboxylic acids as well as C(sp³)−H functionalization of free cyclopropylmethylamines¹⁸ and lactamization of aliphatic amides,¹⁹ led to a significant increase in yield to 82%. In addition, base optimization revealed that the strong base LiOH·H₂O was compatible with the reaction conditions, affording the corresponding product in 85% yield (Table S2). Interestingly, while all the screened monodentate ligands (L1, L2), MPAA (L3–L5, L7), MPAAM (L6), and MPAthio ligands (L8–L10) favored methyl selectivity, the bidentate pyridones (L11, L12) switched methylene activation, delivering fused lactone products (Table 1, Table S4).

Table 1.

Ligands Screening for the β-C(sp³)–H methyl olefination of cyclic aliphatic acids^a,b

^aConditions: 1a (0.1 mmol), Pd(OAc)₂ (10 mol%), ligand (L) (13 mol%), KF (2 equiv), Ag₂CO₃ (3.5 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 100 °C, 24 h.

^bThe yields were determined by ¹H NMR analysis of the crude product using CH₂Br₂ as the internal standard.

The substrate scope of β-C(sp³)–H methyl olefination of cyclic aliphatic acids was subsequently explored under the optimized reaction conditions (Table 2). A broad range of free carboxylic acids bearing five- (1a), six- (1b), seven- (1c), eight- (1d), eleven- (1e), twelve- (1f), and fifteen-membered (1g) rings were compatible, affording regio γ-lactone products (3a-3g). In addition, good yields were obtained with bicyclic systems such as 3h, 3i and 3j. Substitution at the 4-position of cyclohexane carboxylic acid (1l) was well tolerated, delivering product 3l in moderate yield. Acids containing a heteroatom including oxygen (1k, 1n) and nitrogen (1m) afforded the products (3k, 3m, 3n) in 24 to 64% yield.

Table 2.

Scope of Cyclic Carboxylic Acids for the β-C(sp³)–H Methyl Olefination of cyclic aliphatic acids^a,b

^aConditions: 1a (0.1 mmol), benzyl acrylate (0.2 mmol), Pd(OAc)₂ (10 mol%), ligand (L9) (13 mol%), LiOH·H₂O (2 equiv), Ag₂CO₃ (3.5 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 100 °C, 24 h.

^bIsolated yields.

^cL7 as ligand.

Then, the substrate scope of the olefin coupling partners was investigated. As showcased in Table 3, beyond acrylate derivatives (2a–c), various Michael acceptors such as vinyl sulfones (2d and 2e), acrylamides (2f and 2g), vinyl phosphonate (2h), and vinyl ketone (2i) were well tolerated under the optimized conditions, delivering spiro γ-lactones (4a–i) in 30–72% yield. Challenging olefin partners such as dimethyl fumarate (2j) and α-methylene butyrolactone (2k) were effectively incorporated. Moreover, maleimides (4l and 4m) were compatible with the optimized conditions, affording structurally diverse spirocyclic pyrrolidines in moderate yields.

Table 3.

Olefin Scope for the β-C(sp³)–H Methyl Olefination of Cyclic Aliphatic Acids^a,b

^aConditions: 1a (0.1 mmol), olefin (0.2 mmol), Pd(OAc)₂ (10 mol%), ligand (L9) (13 mol%), LiOH·H₂O (2 equiv), Ag₂CO₃ (3.5 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 100 °C, 24 h.

^bIsolated yields.

^cL7 as ligand.

β-C(sp³)−H arylation of carboxylic acids has become a powerful tool for direct C–C bond formation, enabled by Pd catalysis and ligand design, with notable success in linear aliphatic acids and methylene-selective transformations. Despite extensive developments, the selective arylation of methyl C(sp³)–H bonds in cyclic carboxylic acids remains a major challenge, largely due to steric hindrance and the difficulty of forming the corresponding palladacycle intermediates. Thus, we set out to explore our optimized lactonization conditions in β-C(sp³)–H methyl arylation of cyclic aliphatic acids using aryl iodides as coupling partners. The ligand-dependent selectivity in arylation (Table S7) is consistent with that observed in olefination. Monodentate ligands (L1, L2), MPAA (L3–L5, L7), MPAthio (L8–L10), and MPAAM (L6) predominantly favor methyl arylation, while bidentate pyridone ligands (L11, L12) shift the reactivity toward methylene C–H activation, delivering γ-arylation products. A range of cyclic carboxylic acids, including 5-, 7-, and 11-membered ring systems (6a, 6c, 6d), as well as tetrahydro-3-methyl-3-furancarboxylic acid (6b), were found to be compatible, delivering the arylated products in moderate yields. However, cyclohexane carboxylic acid showed low efficiency due to competing methylene-selective side reactions, including γ-arylation¹¹ and diarylation¹³ byproducts. Moreover, a variety of para-substituted aryl iodides bearing electron-withdrawing groups such as trifluoromethyl (6a), ester (6e), nitro (6f), and aldehyde (6g) were well tolerated under the reaction conditions. However, non-substituted aryl iodides (6h) reacted in low yield. Furthermore, the resulting β-methyl-substituted carboxylic acids could be transformed into pharmaceutically important spiro-3,4-dihydrocoumarins (6i and 6k), which are widely present in nature and exhibit a broad of biological activity,²⁰ including anti-herpetic, anti-inflammatory, antioxidant, anti-aging, and anti-cancer properties.²¹ In addition, the corresponding spiro ketone (6j) was obtained via a Friedel–Crafts acylation promoted by methanesulfonic acid (Table 4).

Table 4.

Acids and Aryl Iodides Scope for the β-C(sp³)–H Methyl Arylation of Cyclic Aliphatic Acids^a,b

^aConditions: 1a (0.1 mmol), Ar−I (0.2 mmol), Pd(OAc)₂ (10 mol%), ligand (L9) (13 mol%), LiOH·H₂O (3 equiv), Ag₂CO₃ (3.5 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 100 °C, 24 h.

^bIsolated yields.

^cL7 as ligand.

^d6 (0.1 mmol), CF₃COOH (0.5 mL), BF₃·OEt₂ (0.15 mmol), (Bis(trifluoroacetoxy)iodo)benzene (0.15 mmol), 30 °C, 24h.

^e6 (0.1 mmol), methanesulfonic acid (0.2 mL), 120 °C, 1 h.

Site-selective modification of both methyl and methylene C(sp³)–H bonds in cyclic carboxylic acids represents a valuable strategy for diversifying molecular architectures in drug discovery. Building on our β-C(sp³)–H methyl-selective olefination and arylation strategies, and in combination with previously reported methylene-selective functionalization, we developed a sequential approach that enables dual functionalization at the methyl and methylene positions, allowing for the streamlined synthesis of structurally complex and highly functionalized cyclic acid derivatives. We introduced aryl and heterocyclic substituents at the γ-position of cyclohexane (7c) and cyclopentane (7a, 7b) through a transannular γ-C(sp³)–H arylation enabled by a bidentate quinuclidine–pyridone ligand. Subsequent β-C(sp³)–H methyl-selective olefination, promoted by an MPAThio ligand, afforded γ-substituted spiro-lactones (8a-8c) with high efficiency. In addition, by altering the reaction sequence, we first achieved β-C(sp³)–H methyl-selective arylation, followed by a range of transformations including γ-arylation, β,γ-dehydrogenation/olefination, and [2+2] benzannulation. This strategy enabled access to structurally diverse products such as bifunctionalized cyclopentane acids (8d), benzocyclobutene (BCB) derivatives (8e), unsaturated acids (8f), and unsaturated fused lactones (8g) (Scheme 2).

Scheme 2. Ligand-Control Site-selective Sequential Functionalization of Carboxylic Acids via β-C(sp³)–H Methyl and Methylene Activation.

Conditions: ^a1 (0.1 mmol), (hetero)aryliodide (2 equiv), PdCl₂(PPh₃)₂ (10 mol%), ligand (10 mol%), Cs₂CO₃ (1.5 equiv), Ag₂CO₃ (1.5 equiv), HFIP (0.6 mL), 60 °C, 24 h. ^b7 (0.1 mmol), olefin (0.2 mmol), Pd(OAc)₂ (10 mol%), ligand (L9) (13 mol%), LiOH·H₂O (2 equiv), Ag₂CO₃ (3.5 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 100 °C, 24 h. ^c1 (0.1 mmol), Ar−I (0.2 mmol), Pd(OAc)₂ (10 mol%), ligand (L9) (13 mol%), LiOH·H₂O (3 equiv), Ag₂CO₃ (3.5 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 100 °C, 24 h. ^e6e (0.1 mmol), dihaloarene (0.2 mmol), Pd(CH₃CN)₄(BF₄)₂ (10 mol%), ligand (13 mol%), K₂CO₃ (2.5 equiv), Ag₂CO₃ (2 equiv), HFIP (1.0 mL), 100 °C, 20 h. ^f6a (0.1 mmol), Pd(OAc)₂ (6 mol%), ligand (10 mol%), NaTFA (0.75 equiv), Ag₂CO₃ (2 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 70 °C, 16 h. ^g6c (0.1 mmol), Pd(OAc)₂ (10 mol%), ligand (13 mol%), olefin (0.2 mmol), KF (2 equiv), Ag₂CO₃ (2 equiv), HFIP/MeCN (1.0 mL/0.1 mL), 100°C, 16 h.

To elucidate ligand effects on methyl and methylene C–H activation in cyclic carboxylic acids, H/D exchange experiments (Scheme S1) and computational analyses (Scheme S3 and Scheme S4) were conducted with 1-methylcyclohexane carboxylic acid as the model substrate comparing ligands L7 and L9 with pyridone-based L12, previously employed for selective methylene C(sp³)–H olefination.⁹ H/D exchange experiments further showed that ligand L12 gives 2.4:1 ratio of methyl-to-methylene deuteration, whereas L7 and L9 exhibit significantly higher ratios (up to 6:1), indicating that these ligands markedly enhance selectivity at the CMD step. Additional H/D exchange experiments performed in the presence of coupling partner further support that CMD process is likely irreversible (see Supporting Information, Section 2.2). Computational studies revealed that the transition states for methyl and methylene C–H activation correspond to distinct orientations of the carboxylic acid group (Scheme S3, A). CMD process is likely irreversible and may arise from the formation of more stable intermediates (Scheme S4, L9). The calculated methyl-to-methylene activation ratios and relative reactivity are consistent with the experimental H/D exchange results. Notably, for ligands L7 and L9, although the overall activation barriers for C–H activation are increased, the energy difference between methyl and methylene C–H activation is significantly reduced (Scheme S3). These results suggest that methyl selectivity may arises from the ability of the ligands to effectively diminish the intrinsic reactivity difference between methyl and methylene C–H bonds.

3. Conclusion

In summary, we developed a ligand-controlled strategy for selective β-C(sp³)–H methyl olefination and arylation of cyclic carboxylic acids. This C–H functionalization method opens a new avenue for preparing spiro-carbocycles including pharmaceutically relevant spiro-3,4-dihydrocoumarins and spirocyclic ketones. Computational and deuterium studies indicate that MPAA and MPAThio ligands override the preference for methylene C–H activation observed with bidentate pyridone ligands in cyclic carboxylic acids by selectively promoting methyl C–H activation. Sequential methyl and methylene C–H functionalization further expands carbocycle diversity, offering a powerful tool for rapid synthesis of complex and pharmaceutically relevant scaffolds.

4. Experimental Section

The general procedure for the selective methyl β-C(sp³)–H olefination and arylation of cyclic carboxylic acids reaction was conducted as follows. A sealed reaction tube equipped with a magnetic stir bar was charged with Pd(OAc)₂ (2.2 mg, 10 mol %), ligand (13 mol %), the corresponding carboxylic acid substrate (0.1 mmol), olefin (0.2 mmol) or aryl iodide (Ar–I, 0.2 mmol), Ag₂CO₃ (83.5 mg, 0.35 mmol), and LiOH·H₂O (8.4 mg, 0.2 mmol for olefination; 12.6 mg, 0.3 mmol for arylation). HFIP (1.0 mL) and MeCN (0.1 mL) were added to the reaction mixture, after which the vial was sealed tightly. The mixture was stirred at 600 rpm and 100 °C for 24 h. Upon completion, the reaction was allowed to cool to room temperature, the mixture was diluted with ethyl acetate and acidified with 0.2 mL of formic acid. The resulting suspension was filtered through a pad of Celite using acetone as the eluent to remove insoluble materials. The filtrate was concentrated under reduced pressure, and the crude residue was purified by preparative TLC (hexane/ethyl acetate = 10:1 to 3:1).

Supplementary Material

Supporting Information. 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).